Wikipedia talk:Verifiability/Archive 65

Archive 60 Archive 63 Archive 64 Archive 65 Archive 66 Archive 67 Archive 70

"A self-sourcing example" addition

Regarding this addition by Ilovetopaint, what is that supposed to mean and how is it a helpful addition? Does it mean that if a person adds an example of a concept or similar to an article, the example must not be a primary source? If so, that is false and this addition will aid WP:Wikilawyering.

And what consensus is Ilovetopaint speaking of for this addition? Flyer22 Reborn (talk) 06:33, 29 January 2017 (UTC)

I see that Ilovetopaint is already using this new policy to make edits like this, this and this. I have reverted the addition to this policy, pending further discussion. Flyer22 Reborn (talk) 06:43, 29 January 2017 (UTC)

The inserted text was almost incomprehensible. Policies should be written in plain English. Moreover, there is no consensus for it. The cited RfC in Archive 63 was strictly about "in popular culture" entries, not entries in general. I also question "A tertiary source is even better, if available." which was added by the RfC closer despite not being mentioned by any of the participants. In fact, policy clearly favors secondary sources over tertiary sources; see WP:PSTS for example. Zerotalk 06:56, 29 January 2017 (UTC)
Sounds like the attempted insertion is already covered in WP:SPS and/or WP:SELFPUB, but without the ambiguities. Jclemens (talk) 07:06, 29 January 2017 (UTC)
I'm also rather concerned that it would preclude most examples from mathematics articles necessary to illustrate a concept, technique, or theorem. Sławomir Biały (talk) 11:46, 29 January 2017 (UTC)
Not at all... any decent mathematical text book will already contain examples that illustrate the concepts, techniques and theorems. These can be used (and cited). There is no need for our editors to "create" their own examples. Blueboar (talk) 13:34, 29 January 2017 (UTC)
But the proposed addition to the guideline mandates that citations must show the significance of those examples, which is an unreasonable standard for examples. Moreover, routine calculations are permitted on Wikipedia, and do not stray into OR: Simple numerical examples, for instance, are generally permitted. Slight modifications to standard examples in textbooks are permissible, etc. To quote User:Pmanderson: "If we are not free to come up with exposition, the space between plagiarism and OR narrows dangerously." Sławomir Biały (talk) 17:28, 29 January 2017 (UTC)
There is more written about the subject at Wikipedia:Self-sourcing examples. I've also written an essay (WP:NAMEDROP). And you're right, the consensus was about "in popular culture" entries. I overlooked that detail and didn't consider the problem it would present. Is this better wording?

Examples in popular culture should have sources that not only establish its verifiability, but also its significance in the context of the article.

This is not about WP:SPS, it's a combination of WP:DETAIL and WP:PROPORTION. The article Garage rock is one affected instance. Here is a typical sentence from the article's overly intricate "Regional scenes" section:

The Fallen Angels, from Washington, DC released the single "Bad Woman" in 1966, then went on to record two albums for Roulette Records.[97]

Nothing further is said about group, like if the song was a major hit, or if they even contributed anything to the genre beyond the mere fact that they existed. The only people who would care about this little-known band are devout obsessives, and so it begs the question, "how does this detail enhance my understanding of the greater topic?". There are a lot of these self-sourcing examples in the article — I think it makes the subject overwhelming and inaccessible to the majority of readers like myself who only want a summary of important facts. I think that we need to acknowledge this specific issue, because it's widespread in almost every article about pop culture and the arts (especially music genres). Compare the above with this passage from Psychedelic rock:

The Kinks and the Yardbirds also incorporated droning guitars to mimic the qualities of the sitar,[44] but the Beatles' "Norwegian Wood" (December 1965) marked the first released recording on which a member of a Western rock group played an Indian instrument.[45] The song is generally credited for sparking a musical craze for the sound of the sitar in the mid-1960s – a trend which would later be associated with the growth of the essence of psychedelic rock.[44]

Obviously much more enlightening than "The Beatles, from Liverpool, England released the song "Norwegian Wood" in 1965, then went on to record five more albums for EMI Records". To ensure clarity, I would support copying the full explanation (seen here) that was written by @BrightRoundCircle:. --Ilovetopaint (talk) 19:31, 29 January 2017 (UTC)

(responding to ping) Attempts to incorporate self-sourcing examples into policy failed because, as mentioned above, people worry about the scope. In order to narrow the scope and address the concerns of both sides, I propose the following RfC that is much more limited in scope but addresses the same issue. BrightRoundCircle (talk) 21:38, 29 January 2017 (UTC)

A self-sourcing example can be a problem, especially when a section has many of them, but I don't agree that we always need to explain the matter beyond simply stating that "[So and so] is an example of [so and so]."
I'm not sure why BrightRoundCircle started the RfC at the WP:NOT talk page, but I will weigh in there and let editors know that the RfC is a result of discussion here. Flyer22 Reborn (talk) 20:06, 30 January 2017 (UTC)
Update: Since BrightRoundCircle is proposing that the addition be added to the WP:NOT policy, I now see why he started the discussion there. Flyer22 Reborn (talk) 20:19, 30 January 2017 (UTC)

Where can I file a complaint regarding people using WP:BURDEN to vandalize edits?

No action

Sorry if this is the wrong page to ask, but if someone persistently removed the sentence "GHB is chemically closely related to GABA" from the GHB page, which is a fact everyone can observe with their own eyes on both pages - yet its important to read if you have not memorized the chemical structure of GABA and can derive that information from the top of your head - something an educated researcher could do and doesn't need to write or explicitly hear in their articles hence its hard to source this precise fart anywhere - hence its still an uncited statement on the GHB page, citing WP:BURDEN as a reason, where can I file a complaint for that vandalism / stupidity? 12:29, 30 January 2017 (UTC)12:29, 30 January 2017 (UTC)C0NPAQ (talk) shoa

  • Insisting that information on Wikipedia be properly sourced is neither stupid nor vandalism. Anyone can challenge an unsourced statement; it's a key mechnism for ensuring the information on Wikipedia is accurate. It should not be put back without a source. Reyk YO! 12:51, 30 January 2017 (UTC)
I would add, why not just supply a citation and move on? If it is hard to find a source, that would seem to emphasize the importance of providing one. DonIago (talk) 15:21, 30 January 2017 (UTC)
Exactly. If it's "a fact everyone can observe" then almost any introductory-level textbook on the subject would mentions this fact, and you should have no trouble citing it. By the way, whether or not two psychoactive compounds are most similar to each other definitely needs a source. BrightRoundCircle (talk) 19:52, 30 January 2017 (UTC)
Well, this is weird. As far as I can see the challenged statement appears to be that gamma-Hydroxybutyric acid is "Chemically closely related" to gamma-Aminobutyric acid, i.e. that a molecule that looks like this is chemically very similar to a molecule that looks like this, and I would say that someone who challenges that is in need of support and direction. If it is being persistently removed then rewarding them for their behaviour by obediently scurrying off and finding a source appears to be a suboptimal reaction to the problem. But I can't see a recent challenge in the article history, so I'm mystified.—S Marshall T/C 20:00, 30 January 2017 (UTC)
So you're saying that if any two molecules look alike in a diagram, editors don't need to source statements about their supposed chemical relation? That's very bad chemistry and bad Wikipedia editing. BrightRoundCircle (talk) 20:08, 30 January 2017 (UTC)
No, I'm saying that an editor who doesn't understand those diagrams shouldn't be editing those articles.—S Marshall T/C 20:31, 30 January 2017 (UTC)
Either the claim is verifiable and can be included on Wikipedia, or it's original research and can't be included. Incidentally there's already a source in the article about the chemical relation between these molecules. Claiming that requiring a source confers not understanding the topic is bad for Wikipedia. BrightRoundCircle (talk) 20:51, 30 January 2017 (UTC)
Yes, there's a lot that's puzzling about this complaint. I'm not claiming that "requiring a source confers not understanding the topic", and I have no idea why you think that as I've said nothing that even suggests it. I'm saying that WP:BURDEN is not for use by Randy from Boise.—S Marshall T/C 21:07, 30 January 2017 (UTC)
When something is puzzling about the complaint, you have to see what's really up. Otherwise you will end with calling Jytdog "Randy from Boise". Staszek Lem (talk) 23:28, 30 January 2017 (UTC)
Isn't the first time I've said that about Jytdog, I assure you.—S Marshall T/C 01:09, 31 January 2017 (UTC)
WP:BURDEN is fine when someone wants a source for a claim, for example that "GHGA is chemically closely related to GABA". Calling someone who asks for a source for that claim "Randy from Boise" seems very straightforwardly saying that asking for a source is "uninformed but relentless" as the essay says, when it's actually Wikipedia policy. BrightRoundCircle (talk) 01:49, 31 January 2017 (UTC)
Welcome to WT:V, BrightRoundCircle. I see you're putting this in black and white terms. I think that after you've been here a while your understanding of this will expand a bit. WP:BURDEN is written like it is for very good reasons. It's meant for use in the most difficult and controversial areas of the encyclopaedia where we need to give the heavy artillery to the skeptics. Unfortunately, that wording also makes it into a gift for griefers and people who take a highly controlling approach to article content. People use it to remove unsourced content that doesn't accord with their own point of view, but leave unsourced content that does. People use it to make stupid, idiotic edits, including in one recent case a reckless, negligent content removal that was a child protection risk. Don't be absolutist about WP:BURDEN. Insist that it's to be used intelligently by competent editors who aren't advancing a point of view.—S Marshall T/C 17:51, 31 January 2017 (UTC)
Sorry, no. Requiring a source for a claim is not "griefing" or POV-pushing. Particularly in this case, the editor complaining about having to provide a source appears to be the one griefing, complete with personal attacks and abusive language. BrightRoundCircle (talk) 19:51, 31 January 2017 (UTC)
Ever clash with an editor and then find your whole watchlist lighting up because they're going through everything you've ever written systematically tagging or removing content? BURDEN absolutely can be used vexatiously or for griefing, and is sometimes used for those purposes, and it's happened to me.—S Marshall T/C 20:08, 31 January 2017 (UTC)
An editor can WP:GAME, WP:POINT, WP:TEND, WP:LAWYER, and be disruptive by exploiting any policy. WP:BURDEN is not "meant for use in the most difficult and controversial areas". It's a core policy meant to be used always and everywhere, an essential part of WP:V and WP:RS. If you couldn't source a statement you added to Wikipedia, then it's very possible it doesn't belong on Wikipedia. That's a core policy. BrightRoundCircle (talk) 20:42, 31 January 2017 (UTC)
Sure, I agree with all that. You said earlier that requiring a source wasn't griefing, and it can be. All I'm saying is that when someone says that BURDEN is being used vexatiously or to advance an agenda, or unwisely by someone who's editing an article they're not competent to understand, then we do need to take those reports seriously rather than giving them dogma about policy.—S Marshall T/C 21:00, 31 January 2017 (UTC)
Since that's not remotely the situation here; since no action is being taken anyway; since there are already at least three other editors who said applying WP:BURDEN is correct in this case and none that say otherwise; since the contributor raising the complaint isn't pursuing it; and since you say you agree that WP:BURDEN is "meant to be used always and everywhere"; I think I better just let it go. BrightRoundCircle (talk) 00:46, 2 February 2017 (UTC)
Just as well, a person who cannot provide a reference in support of statement "GHGA is chemically closely related to GABA" shouldn't be editing these articles either. And a person who doesn't understand these diagrams is a valid reader and may pose a valid challenge for clarification. For starters, "chemically closely related" is a WP:VAGUE statement and unless it is completely trivial, it must be explained or at least a reference must be provided which explains how exactly they are related. Staszek Lem (talk) 20:54, 30 January 2017 (UTC)

Primary sources: verifiability vs. reproducibility

Suggestion withdrawn. Staszek Lem (talk) 19:53, 7 February 2017 (UTC)

Quite often I see somebody inserts a sentence like this: "A 2016 survey of 236 college students suggested that females drink 17% more Coke than males" (not really :-)).

Our WP:RS says " Although specific facts may be taken from primary sources, secondary sources that present the same material are preferred". WP:PSTS further cautions, but basically covers WP:SYNTH pitfalls.

At the same time, findings of this type often become a subject of hype in the media, not in the least fueled by strife for grant-generating publicity, and formally one may find lots of secondary sources which just popularize the finding rather than corroborate it.

IMO our verifiability policy is weak in this respect.

Part of the problem is covered by out stance WP:EXTRAORDINARY. However what about mundane findings I cited?

Therefore I would suggest to enhance our WP:V in an appropriate place as follows:

any kind of new findings must be cited from secondary sources which (a) confirm reproducibility of the finding (to satisfy WP:V) and (b) describe the significance of the impact on the area of the study (to satisfy WP:UNDUE). Any primary source cited are for the completeness of the reference only.

My concern comes from recent major scientific misconduct scandals in psychology and sociology, which are reportedly plagued with reproducibility problems (if I read the secondary sources correctly). Staszek Lem (talk) 20:34, 2 February 2017 (UTC)

This is addressed in WP:SCHOLARSHIP: "Isolated studies are usually considered tentative and may change in the light of further academic research. If the isolated study is a primary source, it should generally not be used if there are secondary sources that cover the same content. The reliability of a single study depends on the field. Avoid undue weight when using single studies in such fields. Studies relating to complex and abstruse fields, such as medicine, are less definitive and should be avoided. Secondary sources, such as meta-analyses, textbooks, and scholarly review articles are preferred when available, so as to provide proper context." BrightRoundCircle (talk) 23:23, 2 February 2017 (UTC)
I think that that also helps helps with another common problem which is where there multiple studies and the the Wiki editor cherrypicks the one(s) that align with their POV. North8000 (talk) 23:55, 2 February 2017 (UTC)
No this is not addressed. You didn't read my post carefully. The phrasing " if there are secondary sources" is the "weak link". First, how does one know if "there are 2ndary so". Second, I am talking about just published research, when there is no 2ndary or all 2ndary are just publicity hype, with no real peer confirmation.
Once again, my major idea is to incorporate the concept of reproducibility into our criteria for sci citations. Staszek Lem (talk) 01:14, 3 February 2017 (UTC)
In WP:MEDRS and WP:SCIRS, news coverage of a primary source (even in the news section of a scientific journal) is not considered a reliable secondary source, and rightly so. Someguy1221 (talk) 01:44, 3 February 2017 (UTC)
Yes, it's addressed. As Someguy1221 says, news reports on the study don't make it "not isolated". In order for a study to be "not isolated" it has to be cited in sources "such as meta-analyses, textbooks, and scholarly review articles" or, as you say, other studies that reproduce it. That means that even if you have 100 secondary news sources reporting on a study, if the study is not scholarly reviewed, cited by other researches (such as in works that reproduce the study), or broadly incorporated into textbooks, then it's still "isolated". BrightRoundCircle (talk) 02:23, 3 February 2017 (UTC)
Please cite the policy clause in which it is addressed. The one cited by Someguy1221 is with a loophole I explained. I would suggest a stronger formulation:
Avoid using primary sources for studies which were not critically evaluated by peers in secondary sources.
Staszek Lem (talk) 18:44, 3 February 2017 (UTC)
I quoted it above. BrightRoundCircle (talk) 23:40, 3 February 2017 (UTC)
No you didn't. You keep ignoring my explanation that the text you cited contains a loophole "if there are secondary sources". My proposal is to remove this loophole: secondary sources are not "if", they should be "a must". Staszek Lem (talk) 18:13, 6 February 2017 (UTC)
Wikipedia is not a bureaucracy; example below. BrightRoundCircle (talk) 18:21, 6 February 2017 (UTC)

Keep in mind that this policy is supposed to generally guide all topic areas in WP, hence revising it to address a problem in a specific area is often not appropriate, at least not without checking possible side effects on other topic areas. Special requirements or problems in the specific need to be addressed in special policy for that specific area (like for instance the mentioned WP:MEDRS).--Kmhkmh (talk) 02:50, 3 February 2017 (UTC)

I am not speaking about any "specific area", unless you think that scientific research is a very narrow area not worth special treatment. Staszek Lem (talk) 18:44, 3 February 2017 (UTC)
You seem to be misreading what i'm saying, i'm not saying special areas don't deserve a special treatment in policy, just not in general formutalions in this policy which is supposed to apply to all areas, including those where requiring reproducanility makes little sense.--Kmhkmh (talk) 11:59, 5 February 2017 (UTC)

@Staszek Lem: Reproducibility (in a scientific sense) is imho something that needs to happen outside WP and hence cannot really a criteria for WP.--Kmhkmh (talk) 02:54, 3 February 2017 (UTC)

I don't see the "hence". Reliable and unreliable sources also happen outside WP. My point is that isolated primary sources which report results not yet known to be reproducible are not reliable data for the purpose of wikipedia. In my olden times, a finding was reported only after a series of experiments performed, carefully designed for varying conditions, or after a series of observations. Now, one survey and bamm! a noisy article with far reaching conclusions. Staszek Lem (talk) 18:44, 3 February 2017 (UTC)
As I already stated above there are other policies to deal with problematic one time surveys, other than that we leave it to reputable external (secondary) sources to decide whether a particular one time survey is acceptable or not. From my perspective you are dragging a process that belongs outside WP into WP. WP is not the arbiter on scientific inquiry/criteria/procedure and is not setting up its "own" scientific standards, instead it merely reports and compiles what reputable external (secondary) scientific sources state. And those external sources (and the process in the scientific community) need to consider reproducability and not us.--Kmhkmh (talk) 11:59, 5 February 2017 (UTC)
This seems very specialized. First it covers the narrow case where all of the following are present: Scientific/research type text, scientific/research type source, no secondary sources available. Then within this, it is talking about one attribute (reproduc-ability)amongst many, and usage of this it requires Wikipedians to determine whether or not it is reproducable. And the latter has two difficult possibilities. One is for Wikipedians to determine that there has been no corroborating study (unless this is put on the poster in a "burden" type manner.) The second is finding a corraborating study and Wikipedians determining that it is such. North8000 (talk) 19:11, 3 February 2017 (UTC)
I see what you mean under "specialized". In this sense, yes. But the problem is I see this happening this all the time. And to solve it, no Wikipedians do not have to determine reproducibility all by themselves. All they have to do is to provide a secondary source. Staszek Lem (talk) 23:25, 3 February 2017 (UTC)

Seems to me that reproducibility is a bit of a red-herring. In the first place, while reproducibility is the gold standard of a study (which is to say, of obtaining a certain result in a certain way), there may be other ways of confirming a result. Where a result has been obtained by an experienced, reputable team, with documented and carefully scrutinized methodology, it is quite often the case that everyone in the field is so confident of the result that no reproduction is attempted. Strictly speaking, a result is not "known to be reproducible" until it is actually reproduced, but in many cases that is of the same order as "the fight is not always to the strong, nor the race to the swift, but that's the way to bet." On the other hand, there are results so lacking in credibility that no one is interesting in disproving them. It is quite possibile to have a pair of papers, with diametrically opposite results, neither of which has been reproduced, but one is accepted by the scientific mainstream, and the other is not. In such cases lack of reproducibility in itself is not an adequate indicator of the mainstream view, and an unuseful requirement.

There is also a big difference between "not reproduced because not attempted" (for any reason), and "not reproduced because attempted and failed", neither of which is necessarily a refutation (or disproof). (Failure to reproduce is often not the fault of the original study.) In evaluating contending results editors certainly should consider not just whether a result was reproduced, but also whether that was even attempted, and whether there are other results that corroborate, or discorroborate, either result. And most certainly consider any authoritative secondary sources, such as review articles. But even those can be misleading, such as when different factions have contending review articles.

In the end I don't know that any hard and fast rule can be made, let alone ought to be a strict requirement. While specific statements can be verified in sources, determination of a viewpoint requires a more considered approach. Mere reproducibility, or lack of, is not a useful litmus test. ~ J. Johnson (JJ) (talk) 23:19, 3 February 2017 (UTC)

"Contending review articles" is good, i.e., the result has some publicity and we can report it along with saying that the peer opinions are mixed. Basically, if no peer opinion then WP:UNDUE.Staszek Lem (talk) 23:31, 3 February 2017 (UTC)
OK. I got your point: "reproducibility" is not the only criterion of feasibility. Let's replace it with "reproducibility or other kind of corroboration". Still I'd like to have a way to fend off random hot-plate factoids from primary sources (an example I started with) by fresh post-grads to litter wikipedia. And stricter demand for secondary sources is IMO a solution. Staszek Lem (talk) 23:31, 3 February 2017 (UTC)
Already addressed in WP:SCHOLARSHIP. BrightRoundCircle (talk) 23:42, 3 February 2017 (UTC)
@BrightRoundCircle: My point is WP:SCHOLARSHIP is weak. It has a loophole "... if there are secondary sources". The same loophole is in WP:RS: "secondary sources that present the same material are preferred". I.e. Current policies do not fend off factoids from articles which nobody even read yet. Staszek Lem (talk) 18:00, 6 February 2017 (UTC)
An editor adds information from an isolated study that no one has reproduced, with 100 secondary sources reporting on it. You think it's inappropriate for an encyclopedia because it's an isolated study that no one has reproduced. You remove the information and put "WP:SCHOLARSHIP" in the edit summary. Your removal is reverted; on the talk page the other editor has to explain why the study isn't isolated. If they can't, the information is removed again until there's consensus the study isn't isolated. As long as this process can take place, there's no loophole. There's no need to specify "must" in every policy and guideline, Wikipedia is not a bureaucracy. BrightRoundCircle (talk) 18:21, 6 February 2017 (UTC)
@BrightRoundCircle: Forget "reproducibility". I am regretting I used this term already. It is like a red rag for a bull now. For the nth time, if secondary sources report it then it is OK. Heck, we even have article about pseudoscience here. What is not OK is when there are no secondary sources at all. And removal of "isolated results" is not mandated by the language of WP:SCHOLARSHIP. My suggestion (minus reproducibility) clarifies the language to mandate this. Staszek Lem (talk) 18:28, 6 February 2017 (UTC)
Answered below. BrightRoundCircle (talk) 18:34, 6 February 2017 (UTC)
I agree that WP:SCHOLARSHIP is somewhat relevant, but it does so in the context of WP:Identifying reliable sources, and not on how to weigh sources in assessing a viewpoint. And it says nothing at all about reproducibility or its significance.
And I agree that "hot-plate factoids" (I kind of like that) should not be included simply because it can be verified that someone has said as much. (Especially in the popular media, but in primary sources as well.) But! reproducibility is not how we corroborate a result – that would be original research. We corroborate by finding other sources. Which might affirm a result by reproducing it. Or getting the same result by a totally different method. Mainstream opinion is not always based on reproducibility; to make that a requirement would just provide another basis for doing mischief. ~ J. Johnson (JJ) (talk) 00:06, 5 February 2017 (UTC)
+1 I completely agree.--Kmhkmh (talk) 12:04, 5 February 2017 (UTC)
OK. Forget "reproducibility". I keep repeating that the current policies are formulated with loopholes. (see above). And I an suggesting how to remove them and thus prevent the "factoids". Staszek Lem (talk) 18:17, 6 February 2017 (UTC)
An editor adds sourced information that you think is inappropriate for an encyclopedia. You remove the information and explain why in the edit summary using the proper Wikipedia policy. Your removal is reverted; on the talk page the other editor has to explain why the source and information are appropriate. If they can't, the information is removed again until there's consensus it's appropriate. As long as this process can take place, there's no loophole. There's no need to specify "must" in every policy and guideline, Wikipedia is not a bureaucracy. BrightRoundCircle (talk) 18:34, 6 February 2017 (UTC)
An editor adds a factoid not covered in secondary sources. You remove it saying isolated facts are not allowed in wikipedia. The original poster says that "isolated facts" are not forbidden in wikipedia. You say yes the are. He says "please cite the exact clause which forbids "isolated facts". You point to WP:SCHOLARSHIP. He says not a single sentence of WP:SCHOLARSHIP forbids "isolated facts". You say "yes it forbids, see my quote above". He says: "I have already 'seen above' and I see what while is does dislike 'isolated facts', still it does not forbid them, only gives a preference to secondary sources". Your colleague pops up and says "yes isolated facts are bad". The original poster asks "why on Earth a fact cited from a peer-reviewed magazine is bad?" <Rinse, repeat...> Eventually, seasoned editors bully the new contributor off wikipedia, because Wikipedia is not a bureaucracy. Sure thing, colleague, you win. Maybe. But after much wasted time. Staszek Lem (talk) 20:09, 6 February 2017 (UTC)
Reply below. BrightRoundCircle (talk) 03:46, 7 February 2017 (UTC)
  • Is there a specific edit or situation that we're discussing here, or is all this theoretical?—S Marshall T/C 18:40, 6 February 2017 (UTC)
    • See the very top of the section. Haven't you seen these kinds of edits? Staszek Lem (talk) 20:09, 6 February 2017 (UTC)
      • Well, yes, I have, a few times, over the years. I'm not sure the verifiability policy is the place to address them. I've always thought the real problem is overemphasizing recent reports that haven't had enough peer review, and I've always seen it as more of a NOR issue ---- or, ideally, an editor in need of some mentoring, support and direction. What triggered this thread? Is there a recent instance we can look at?—S Marshall T/C 20:19, 6 February 2017 (UTC)
        • I started this thread here because I (probably erroneously) linked the problem with the notion of reproducibility. I don't think it is NOR. In the course of this talk I came to the conclusion that the simplest way to solve the problem is to close the #loophole in the formulations of WP:RS and WP:SCHOLARSHIP. @S Marshall: As for "what triggered this thread", there was no drama; I reverted here, uncontested, but I still had second thoughts that I did that "not exactly by the book". Staszek Lem (talk) 20:57, 6 February 2017 (UTC)

Sounds like deliberately missing the point. The policy WP:SCHOLARSHIP explicitly states "Avoid undue weight when using single studies". If they try to force the inclusion of the isolated study because the policy uses the wording "avoid" instead of "YOU MUST NOT UNDER ANY CIRCUMSTANCES!!!" then they are not here to build an encyclopedia. BrightRoundCircle (talk) 03:46, 7 February 2017 (UTC)

  • You're overdramatising and hard lining. Nobody tried to force anything. An editor added a source they felt was relevant; another editor reverted because they felt it was better left out; and the matter rested there. I don't see anything of concern there at all, and I don't see a need to edit the policy in response to what happened.—S Marshall T/C 17:51, 7 February 2017 (UTC)
  • OK. You convinced me that "avoid undue weight" may work, despite the fact that this phrasing is vague. I am withdrawing the suggestion until a real drama happens, although I still believe that a bit stricter wording might be better, including the clarification of the term "isolated studies". Staszek Lem (talk) 19:53, 7 February 2017 (UTC)

Semi-protected edit request on 12 February 2017

Rabbih020 (talk) 11:04, 12 February 2017 (UTC)
  Not done: it's not clear what changes you want to be made. Please mention the specific changes in a "change X to Y" format. DRAGON BOOSTER 12:02, 12 February 2017 (UTC)

Paywall issue

I understand that a paywall may exist for magazines, newspapers, research articles, etc. But in general this information is available to general public with a little extra legwork. But what about citing a private firm which makes money off publishing "industry reports", available only for money? I would say it puts an unnecessary burden on verification, as well as an undue promotion of a business in wikipedia. There are hundreds of such "market analyzers". IMO, if mainstream cites them, we cite the mainstrem, with credits to the private report. Staszek Lem (talk) 00:48, 31 January 2017 (UTC)

We're talking the $x,000-type marketing reports, right?
My take is that if WP:V allows rare sources that may only be in a few locations and would require special arrangements to get to and see ($$$), then these as PAYWALL would be fine. But the question becomes how reliable are those reports are: they are written as marketing tools to project what the market will be in X years, and thus could be seen as unnecessary speculation. I would agree that we'd rather see a reliable third-party say "According to analysis firm ...". PAYWALL here doesn't eliminate those reports outright, but it becomes a matter of vetting authority, editorialism, etc. which is much easier to ascertain if a third-party says the same. --MASEM (t) 00:59, 31 January 2017 (UTC)
I am talking to only about future speculations, but also annual reports. Staszek Lem (talk) 01:13, 31 January 2017 (UTC)
Are you talking about annual reports of public corporations? That some of them are behind a high $ paywall? North8000 (talk) 01:25, 31 January 2017 (UTC)
No, I am talking about "market researchers". Staszek Lem (talk) 02:25, 31 January 2017 (UTC)
I think the primary concern is reliability, not price. If the source is regarded as a sufficiently important "published source with a reputation for fact-checking and accuracy", it shouldn't matter that it's expensive; a sufficiently motivated group could find ways to raise funds for buying a copy, and maybe open a WP:OTRS ticket to verify that a Wikipedian indeed was able to verify the content that supports a relevant point in the article.
If the fact that it's accessed by few people makes you doubt of its reliability, that should be the reason to avoid using it; but if there are good reasons to think it's reliable nevertheless, then it still can be used. Diego (talk) 11:59, 31 January 2017 (UTC)
I do not think it is reasonable to expect editors (or the Wikipedia foundation) to have to pay in order to satisfy WP:V. This strikes me as inherently gameable and problematic. If a source is not held in libraries, university collections, or other such "freely" accessible place, then I do not believe that it meets WP:V, WP:PAYWALL notwithstanding. It seems to me that WP:PAYWALL is intended to cover more conventional kinds of paywalls on sources that are held by libraries and university collections, and that can in principle be accessed without individual editors (or the foundation) being required to spend money to access. Sławomir Biały (talk) 13:16, 31 January 2017 (UTC)
I disagree. While you (or I) may not be willing (or able) to pay to access a PAYWALLED source, as long as some other editor (one with more money) is able to access it, the source is verifiable.
To give an extreme analogy... suppose there is a rare document located in a remote location (say a library in Mongolia). You and I probably can not afford to fly to Ulan Bator to view this rare document ... but our inability to travel does not change the fact that the information in that document is still verifiable. Someone with more resources than you or I can fly to Ulan Bator and view it. The standard is that information be able to be verified ... not that you (or I... or any other specific editor) be able to do so. Blueboar (talk) 14:33, 31 January 2017 (UTC)
Agree with Blueboar. Plus, starting a new condition like this could have a lot of unintended consequences. Just a claim that the paywall is too high could justify knocking out sourced material. North8000 (talk) 15:40, 31 January 2017 (UTC)
Maybe the trick to satisfy Staszek Lem's concerns regarding "market analysis" would be to treat by default this kind of paywalled, rare access sources as self-published sources, at least until we have a review of their content by a different reliable source?
This way we could quote them for the relevant parts, attributing article parts as "industry report by marketing analyst X says that, in 2017..." This, without requiring the readers to access the source to verify that a relevant fact asserted in Wikipedia's voice has been indeed stated by the source. Diego (talk) 17:59, 31 January 2017 (UTC)
Perhaps I am being unclear. Suppose that a publisher were to insert references to a study that is not held in libraries, and for which they charge $10,000 for viewing. I think that this creates an unreasonable burden for any editor to verify the source, and it would be perfectly reasonable to remove this sourced content as unverifiable. Standard paywalls are things that universities and libraries have access to, they are not meant to include gilded secrets of the illuminati that can only be accessed with massive expenditures from volunteer editors. That creates an unreasonable burden for verification, as well as a problematic avenue for revenue generation at editors' expense. (I could claim to have a reliable, published source, but I'll only show it to you for $10,000!) Sławomir Biały (talk) 18:33, 31 January 2017 (UTC)
I think the question still comes down to how many of these for-profit, high-cost reports are actually usable as reliable sources that includes appropriate content for us to write about (under WP:NOT considerations); my take is very few of these are usable because they are overly speculative, and as suggested by Diego Moya, could be taken as SPSs. There are definitely times that other sources will point to these reports, and that is where we reasonably can use the third-party reference as the point of WP:V, assuming the third-party is an RS (so that they aren't deliberately mis-quoting the report). In otherwords, yes, these reports are acceptable under PAYWALL despite their cost but generally fail other parts of V/RS policies or do not provide content acceptable under NOT, on their own. --MASEM (t) 18:38, 31 January 2017 (UTC)
I would agree with that. The question of whether a source is appropriate or not is separate from the issue of accessibility and paywall. Blueboar (talk) 22:03, 31 January 2017 (UTC)

For the types of paywalls the OP is talking about, I'm not sure these can even be accurately described as having ever been published. That alone would disqualify them as reliable sources. Someguy1221 (talk) 22:27, 31 January 2017 (UTC)

Agreed. I've come across, and used, the likes of Gartner "magic quadrant" reports on particular areas of tech, for example, and it'd be to the advantage of a company with a product or service deemed to be in the "innovative/advanced" quadrant to highlight that; but the fact that Gartner reports are essentially self-published and normally accessible only to the relevant industry and their customers without paying $$$ means they can't be classed as WP:RS. BastunĖġáḍβáś₮ŭŃ! 12:46, 2 February 2017 (UTC)
I'm not really happy with this (apparent) consensus assessment, because it appears a slippery slope to conclude that industry reports in general are not WP:RS because they are self-published, and I don't think that's a good outcome of this discussion. It would be better, in my opinion, to exclude sources not held by library or university collections, that are only in private collections (visible for a fee). Sławomir Biały (talk) 13:27, 2 February 2017 (UTC)
Unfortunately, as more and more works go digital-only, that type of requirement becomes problematic too. There are plenty of complete RS-meeting books and works that exist only as for-pay digital-only versions, and we're still at a stage of how public libraries and schools are going to be able handle those works from a copyright standpoint. That's why its better to accept that PAYWALL would allow these types of market reports (as it would be difficult to distinguish from other more legitimate works that had the same for-cost issues), and instead recognize that these market reports would rarely meet the RS requirements, because they tend to lack editorial oversights, and that they are typically only useful for long-range speculation. Whereas for-cost legit RS still meet RS requirements. It's a much cleaner answer that doesn't require any policy changes. --MASEM (t) 14:36, 2 February 2017 (UTC)
What about the requirement these paywalls are "breachable" in libraries and colleges? Staszek Lem (talk) 20:41, 2 February 2017 (UTC)
This seems to me an essential feature of verifiability, that sources must "in some sense" be checkable by the public at large, not merely the pecunious elite, private guild members, etc. When a "publisher" creates unreasonable barriers to information access, that seems to argue against the source being "published" in the sense of this guideline. Content should in principle be checkable by the general public, equipped with a suite of library resources, but not necessarily with a grant to purchase the requisite materials from private entities. Sławomir Biały (talk) 21:43, 2 February 2017 (UTC)
I'm not convinced. If you want the Gartner reports (an example mentioned above), then I'm certain that you can get the information by asking around on wiki. In a community this big, and this tech-focused, multiple editors will have access to some of them at work.
The free-or-free-from-a-library focus is not helpful. If that's really the limit of what could be verified, then the use of {{cite sign}} needs to be banned and the template deleted. Why? Because no matter how fabulous your library is, it can't send you to the physical location of the sign.
I agree that material must be accessible to the general public. However, "the general public" doesn't mean "people who are not able and/or willing to spend their own money". Sources available only to private guild members (or, to give an actual example that's come up before, members of a particular religion) are not "published". But a source that's only available to people who are willing to spend money on it is still available to the general public. WhatamIdoing (talk) 04:46, 16 February 2017 (UTC)

Translating articles from other Wikipedias or copying from other articles & responsibility for sources

Looking at Wikipedia:Translation there is nothing about checking sources. I believe that we should expect editors translating articles or copying material from our articles to make sure that the sources meet our criteria and back the statement that they source. I've certainly found instances where badly sourced material has been moved from one of our articles to another with no check. Should we add something here and then to Wikipedia:Translation about this? Doug — Preceding unsigned comment added by Doug Weller (talkcontribs) 15:50, 12 February 2017 (UTC)

In fact, the issue is a bit wider. It so happens I often consult Lithuanian wikipedia for bios and I noticed their WP:BIO criteria are way less strict. I can see this is reasonable for a local wikipedia for a small country. However this leads to the problem with translations similar to yours.
Now, answering exactly yours, WP:RFT does contain an advice something to this end: "If portions of an article appear to be low-quality or unverifiable, use your judgment and do not translate those portions." However I agree that the corresponding section of WP:RFT may benefit from a stronger language, especially as a reminder for non-en:wikipedians accustomed to weaker policies. How about something like this:
The translator is advised to ensure the translated text follows core content policies of English language wikipedia, as well as our notability criteria. Use your judgement to either omit inadequate portions or tag them (or the whole articles) appropriately, for other wikipedians to improve the text.
If others agree this is an issue, I suggest to move the discussion to Wikipedia talk:Translation, with notifications is other core policies talk pages. Staszek Lem (talk) 19:24, 14 February 2017 (UTC)
  • Anyone who's interested in the verifiability of translations is invited, and by "invited" I mean "begged", to help me go through the 3,583 articles listed here, because it's a big job and I'd like to finish before I retire.—S Marshall T/C 22:15, 15 February 2017 (UTC)
The usual request along these lines is "would somebody please require all translators to personally obtain and personally inspect every single source before translating an article". The answer is "no". (Ditto for splits and merges, which also attract this request.) WhatamIdoing (talk) 04:15, 16 February 2017 (UTC)

I tried do answer the call of S Marshall, clicked a random article, Carltheo Zeitschel, and after wasting several hours I swore to myself never ever touch a machine translation with a 7-foot pole, other than to convert it into a stub. Staszek Lem (talk) 01:41, 17 February 2017 (UTC)

Can there be an exception case to allow blue links as references on lists?

See Talk:List of symphony orchestras in the United StatesUnscintillating (talk) 23:17, 11 February 2017 (UTC)

Absolutely not, in the sense that if it is deemed WP:MINREF material, links to other Wikipedia articles cannot be used to verify content. We use reliable sources for verification, and Wikipedia articles are not that. If content is challenged (which the talkpage discussion obviously does), inline references to reliable sources that directly support the claims are necessary. – Finnusertop (talkcontribs) 23:31, 11 February 2017 (UTC)
There is no consensus regarding the amount of references lists need. People already mentioned WP:LISTVERIFY and WP:MINREF. This issue needs to be hashed out among the broader Wikipedia community. I'm all in favor of inline references for everything, but there's no consensus. BrightRoundCircle (talk) 23:44, 11 February 2017 (UTC)

While there is no Category:Symphony orchestras in the United States or other subdivision by country in Category:Symphony orchestras, how is such a list as List of symphony orchestras in the United States different from a category? Both list only articles and make the consequent assertion that those articles belong to that group (i.e., that they are both symphony orchestras and that they are in the United States). In the case of any category, one must click through to the article to verify whether it has been added to that group correctly or incorrectly. Why is it more problematic to have to do that on a list, particularly where the alternative is to possibly include hundreds of citations for such basic facts about a subject? If the orchestra is notable, then just about every source about it that exists is going to verify what it is and where it is. Other lists the answer will come out differently, particularly where they annotate or explain, or where they are organized around more complicated or arguably subjective characterizations, but this list is such an elementary one. Pick your battles and consider what we'd actually be accomplishing by copying from its article a citation to accompany every link to that article everywhere it appears. If at the end of the day "conceptual consistency" is the only answer, then it hasn't been a day well spent and you've bloated article code needlessly. postdlf (talk) 01:09, 12 February 2017 (UTC)

  • The issue at Talk:List of symphony orchestras in the United States is not resolved.
    I have explained the following: Wikipedia is not a reliable source, verifiability requires reliable sources, and mainspace list articles require verifiability.
    The question I asked here, is there an exception case, has been answered, no, there are no exceptions.  Postdlf chooses to impose his personal will on the article in question in defiance of the consensus here, thus WP:V is weakened, and editors don't know what will happen at the next List article where references are removed.  Unscintillating (talk) 02:13, 14 February 2017 (UTC)
no, there are no exceptions - there is no community-wide consensus in that regard—see WP:LISTVERIFY "obviously appropriate material". My own preference is for every list item to be referenced, but this is not the consensus. You will either need to resolve that on the article's talk page or bring it to the attention of the wider Wikipedia community. BrightRoundCircle (talk) 06:19, 14 February 2017 (UTC)
WP:LISTVERIFY yields to the sourcing policy, it does not set policy.  If some blue links are now verifiable, where was this decided and what defines the new boundary?  And even if there is potential to build a consensus to move the boundary for symphony orchestras, WP:LISTVERIFY explicitly allows further citations when it states, "When...editors choose to name more sources than strictly required, then either general references or inline citations may be used."  Where is there any case for removing inline citations on the grounds that "I don't like them" or "They look like external links to me".  I don't see anything like that at WT:Verifiability/Archive_61#Can_lists_of_notable_people_be_removed_under_WP:BURDEN?  Unscintillating (talk) 02:50, 15 February 2017 (UTC)
The reasons you were given weren't very good, but that doesn't mean there are no exceptions. I strongly suggest you solve this on the talk page or go through more broad dispute resolution, namely WP:DRR/DRN, or the more appropriate WP:RFC since your issue has to do with core policy. Then you will have your consensus. BrightRoundCircle (talk) 04:22, 15 February 2017 (UTC)
This is clearly the case of WP:V in its very nutshell, which basically saying that info must be verifiable. Quotations are but the means to make verifiability easier for the reader, especially when the statement may be challenged. Now, as to the lists. In our case, how challengeable is a statement that something is a "symphony orchestra in the United States"? Of course, a cn-troll may challenge everything. However if in an extreme case when the blue link does not give a satisfactory answer, one can slap a rare {{cn}}. Otherwise IMO is is pointless to have a page with 80% of text being rather trivial references.
At the same time I may easily imagine lists for which to verify the list criterion from the blue link will require much more effort compared to an external link directly to the info in question. Staszek Lem (talk) 18:57, 14 February 2017 (UTC)
I've asked the OP on the talk page to comply with WP:V's explicit statement that CHALLENGEs need to be because of a sincere concern about the unverifiability of the material. It's an experienced editor, not some troll, and the policy's fairly clear about that limit. That step may move us past the my-policy-trumps-your-common-sense stuff, and let people get back to improving the article. WhatamIdoing (talk) 04:13, 16 February 2017 (UTC)
Quick update: This is not a CHALLENGE; this is not MINREF material; everyone seems to agree that it is extremely un-LIKLEY that anyone would be concerned about the verifiability of the list entries. It seems to basically be a difference of opinion about whether inline sources should be added even though (definitely) not required and (maybe) not very useful. WhatamIdoing (talk) 20:52, 17 February 2017 (UTC)

Exemplary footnote

I think we all have a lot to learnCnbrb (talk) 01:58, 16 February 2017 (UTC)

What the heck is that all about? Drumming up hits on someone's tweet? ~ J. Johnson (JJ) (talk) 23:33, 17 February 2017 (UTC)
Well, <ref>I dreamed this up</ref> is kinda applicable in many places in WP. Staszek Lem (talk) 00:41, 18 February 2017 (UTC)

Reliability of social media sources representing companies

Conclusion: if social media are reliably associated with official spokespersons for the company, then see WP:SELFPUB. Staszek Lem (talk) 19:53, 7 February 2017 (UTC)

Hi all,

How do I establish that certain people, websites etc are, in fact, reliable sources within a specific context?

Specifically in regards to this passage:

   Anyone can create a personal web page or publish their own book, and also claim to be an expert in a certain field. For that reason, self-published media, such as books, patents, newsletters, personal websites, open wikis, personal or group blogs (as distinguished from newsblogs, above), content farms, Internet forum postings, and social media postings, are largely not acceptable as sources.

I accept that generally speaking, forums, social media etc are unreliable. But there needs to be a method of establishing that certain sources are acceptable. For example, a link to a specific post in a facebook thread where the Yarra Valley Railway claims that they will acquire certain vehicles ex South Gippsland Railway. It's an official statement from the YVR, so why is it not considered reliable? It's fairly clear if you look at any of the other posts there that the page is legitimate.

Anothersignalman (talk) 13:53, 26 January 2017 (UTC)

With social media, I would be concerned of link rot, among other (perhaps more substantial) issues. If press releases are issued via social media, and referenced in Wikipedia, then archive links should be strongly encouraged, if not required. Sławomir Biały (talk) 13:58, 26 January 2017 (UTC)
Link rot is where the link expires and no longer functions? Good point then. Need to test if I can archive the above via wayback machine. If not, what's another solution? Anothersignalman (talk) 14:14, 26 January 2017 (UTC)
Wayback Machine doesn't seem to be compatible with facebook. On the other hand, the content of that thread is likely to be printed in, say, Newsrail, in the next few months; so the link doesn't need to be permanent. What are the other "more substantial" issues you mentioned? Anothersignalman (talk) 14:23, 26 January 2017 (UTC)
And here you go, a perfect illustration of ephemerality of social media: the Ayarra link cited above says "recently acquired" Gippsland, not "will acquire", and it is a photo of a poster, asking for donations, not an "official document". An official statement on social media may come and be withdrawn next day; why bother and archive it WP:RECENTISM. If an intention is serious, there must be serious documents, at least at company's official website, otherwise just WP:UNDUE. (And if a company's only official website is the facebook page, I would seriously doubt whether the company deserves a wikipedia article at all.) Staszek Lem (talk) 18:00, 26 January 2017 (UTC)
Re "recently acquired" vs "will acquire" - my fault, bad choice of words. YVR has acquired the vehicles legally but not yet physically - the transfer still needs to happen. There's a few thousand tonnes to move and they don't have a connecting track, so it might take a while, but in the meantime it still makes sense to record that those items are now under the control of, and the responsibility of, YVR not SGR. Even if the vehicles stay at the former SGR site, they'll still be allocated to YVR. In any case, the URL I provided above is supposed to link to a single post about halfway through the thread, starting with "VFBF 7", which is one of the 11 (not 9) vehicles to be transferred. The group does have a website, but it doesn't list their assets. Anothersignalman (talk) 18:48, 26 January 2017 (UTC)
That a company does not list its assets makes sense due to the fluidity of this information, and hence of its dubious encyclopedicity. Wikipedia is not in the business of data mining; it is an encyclopedia. If assets are not officially listed by the company, this info is probably not important, and collecting it via indirect means, such as the poster you cited, provides no guarantee of completeness or correctness of this info, hence, it does not belong to wikipedia IMO. Staszek Lem (talk) 19:11, 26 January 2017 (UTC)
Agree not important to articles like Yarra Valley Railway, but what about articles like Diesel Electric railmotor (VR), which includes the vehicle top-right of the poster, or VBBX23 (listed below VFBF7), included in Victorian Railways box vans#Large Bogie Steel Boxvans? (VFBF7 would go on a future article, Victorian Railways flat wagons, which I haven't had time to write yet.) Anothersignalman (talk) 20:16, 26 January 2017 (UTC)
What is the question? Please be specific; "what about" is not. Also, I suggest you to consult Wikipedia:WikiProject Trains. I am sure, people there must have reasonable traditions and guidelines how to write about railroads and trains. Staszek Lem (talk) 22:47, 26 January 2017 (UTC)
Sorry, the above is in reference to the single facebook post by YVR, in the above thread, which identifies eight of the 11 vehicles that they are now responsible for. It's the post that starts with "VFBF7", and the URL should go directly to it. Anothersignalman (talk) 08:02, 27 January 2017 (UTC)
Back to the original question. The reliability of a source consists of three things: reliability of the publisher, author, and specific document. In our case there is no doubt that Facebook faithfully publishes what the author posted. Further, since the author is on a business-related webpage, we have to consider whether the author is in the capacity to officially represent the business. Further, whether the particular text is an opinion of the person, even if official, or an official statement. Finally whether the particular piece of information from the text contradict the info coming from elsewhere. As a stark example, imagine if we start adding texts into wikipedia articles sourcing them from Trump's tweets. Staszek Lem (talk) 19:24, 26 January 2017 (UTC)
Response to your points - yes, the facebook author by the name of "Yarra Valley Railway" represents the company/group/business (as demonstrated by other posts on the page), and yes, listing the stock (VFBF7 etc) is an official statement, and there is no other information currently available because the transfer of stock is an ongoing event. That information will probably come to light in printed form over the next few years, but in the meantime that would leave existing pages outdated and wrong. (As for your Trump example, yes, it would be perfectly appropriate to include those tweets as sources if the article is, say, Things tweeted by Donald Trump. Context is key, and the discussion as to whether an article should exist or not is a separate issue.) Anothersignalman (talk) 20:16, 26 January 2017 (UTC)
We don't write "Things tweeted" or "things facebooked" articles. As for "ongoing event", see WP:RECENTISM. We are encyclopedia, not a newspaper, nor a replacement for company's website, which does not bother to keep it info up to date. What if during the transfer one engine gets stolen? :-) An encyclopedic article must be reasonably general, not a list of each nook and cranny. But again, this is not the issue of WP:V policy. Staszek Lem (talk) 22:40, 26 January 2017 (UTC)
It's not reasonable to expect YVR (or most other rail heritage groups) to keep their website up to date given that it's a volunteer organisation with limited resources. That's probably why they prefer to publish updates on Facebook -you don't need any coding experience. If one of the engines gets stolen, it'd still be under the control of YVR. If it spontaneously combusts, then I'll update the article to reflect that. Re WP:RECENTISM, my pattern for all the VRLocos pages is to talk about the history of the group of vehicles, any with specific differences affecting modern interpretations, then the units still existing and where they are so that if readers want more information they know who to ask and/or where to go. My focus on the vehicles listed in the YVR article is because, generally speaking, they're the last vehicles of those types. So I think spending one or two paragraphs on each of the Template pages (but not the operator pages, i.e. Yarra Valley Railway) is justified. WP:NOTCATALOG is similar - I'm not listing all the thousands of vehicles, just focusing on the handful that still exist, to assist people who want to do their own research. Anothersignalman (talk) 08:02, 27 January 2017 (UTC)
Mostly agreeable, with one caveat. Since you say YVR is a heritage group, they are supposed to keep their records carefully, especially about things of historical value. Since, as you say, the stock is not thousands, it is really not that hard to update info in their official webpage. Preservation is not just about someones' personal collection to enjoy in private, right? So, if they are serious, they should systematize public info about their assets, rather than let people chase all over their facebook posts. Once again, wikipedia is not in a job of keeping their record straight. That said, you provided reasonable argument that this particular piece of info is reasonably reliable, but, as someone else wrote, this reference is in danger of digital obsolescence. Another problem is that if you had given a link to a centralized list, users could have poked around and see more on the subject, but this link to a random photo gives no additional useful context. Staszek Lem (talk) 17:01, 27 January 2017 (UTC)
Sorry, missed this one previously. None of the heritage groups in Victoria keep perfect records, and the government authorities responsible for keeping track of allocating the government-owned assets frankly have no idea where half the items are or who is responsible for what; in the period 1995-2005 a lot of items changed hands without transfer or publication of records. There was a recent case of a government-owned vehicle being sold by private individuals before anybody realised it was a government-owned, allocated asset. The records are reasonably reliable up to around 1995, and after that everything wound up in a heap so cross-referencing piecemeal information is the least-bad option. Combine that with the limited resources of YVR (and other organisations), and odds are there isn't anybody with the skill set to update the YVR website. As for digital obsolescence, yes, that's a serious risk, but I expect that before it becomes a problem the data will have been replicated in one of the reliable printed references, i.e. Newsrail. I haven't checked recent editions, my collection is a few months out of date but I plan to resolve that in mid-March or April 2017 at the latest. The URL provided for the reference should go to the specific post in the thread, not to the header image. Far from perfect, but again, least-bad option. Anothersignalman (talk) 09:14, 20 February 2017 (UTC)
I think you have a authenticity problem, among other things: no one apparently knows who this Yarra Valley Railroad person is. Mr. or Ms. Railroad (or should we call him/her Yarra) cannot prove he or she is who they claim to be. -- Alanscottwalker (talk) 23:04, 26 January 2017 (UTC)
I thought about that, but it's fairly clear based on the other posts by that account on that facebook page. Anothersignalman (talk) 08:02, 27 January 2017 (UTC)
If it helps, this is the post I'm referring to, with links to where they'd be referenced:
   VFBF 7 - page not yet written
   BP 97
   GY x 3 - which three vehicles still TBA, so that edit would wait until they're identified
   ZF 24
   VBBX 23
   CW17
   RM 55
   MT 200
   Y135
Such speculation, still leaves the authenticity problem. -- Alanscottwalker (talk) 15:13, 27 January 2017 (UTC)
Per WP:VERIFIABILITY, authenticity is a problem when questioned. Our case it is easy to verify that YarraValleyRailway indeed belongs to Yarra Valley Railway. Staszek Lem (talk) 17:01, 27 January 2017 (UTC)
Who is verifying that the account YarraValleyRailway is who it says it is? When someone makes a statement to a reliable source, the reliable source verifies it is who it says is, who is doing that here? Alanscottwalker (talk) 17:50, 27 January 2017 (UTC)
In this case, look through any of the other posts on the YVR facebook page. Anothersignalman (talk) 09:06, 28 January 2017 (UTC)
Facebook is not considered a reliable source, and it cannot be that posts by that account verify posts by that account. Alanscottwalker (talk) 15:27, 28 January 2017 (UTC)
The point of this thread is to figure out why Facebook isn't considered a reliable source. It's easy to verify that the posts are official, because if they weren't the account wouldn't have lasted so long. Anothersignalman (talk) 16:53, 28 January 2017 (UTC)
Facebook has no reputation for fact checking or accuracy. Alanscottwalker (talk) 12:08, 29 January 2017 (UTC)
True, but in this particular case that lack of reputation is easy to overcome because the account has been active for an extended period. And incidentally, does that mean that a blue-tick Twitter account would qualify as a reliable source? Anothersignalman (talk) 12:31, 29 January 2017 (UTC)
To make sure its clear - I am not advocating acceptance of all facebook posts as "reliable sources". I'm advocating this one particular post, by one particular author, based on the reliability and reputation of earlier content by that same author. Anothersignalman (talk) 12:33, 29 January 2017 (UTC)

I don't think it is really warranted to view the Facebook account with such extreme skepticism. Other self-published sources, and even "official websites", might be viewed with similar skepticism, and yet are acceptable in some limited capacity on Wikipedia. Even highly reputable academic presses are not in the business of systematically checking personal identifications to make sure that an author of a submitted manuscript is who they claim to be, as famous pen names show. So unless there is a positive reason for engaging in such skepticism, I do not think it is appropriate, and it may actually be harmful to our project. I would accept social media as an authentic primary source. However, it is still a primary source, and not a very good one. Generally speaking, anything we add to Wikipedia should be based on secondary sources. It's hard to see how minutiae like those things posted on Facebook add anything of lasting encyclopedic value. Sławomir Biały (talk) 12:39, 29 January 2017 (UTC)

Thanks Slawomir Bialy. In this case, tracing the few remaining vehicles of each group allows a reader to plan trips to visit those vehicles if they need to do further research, i.e. measuring components etc. Anothersignalman (talk) 12:59, 29 January 2017 (UTC)
No. It's rather simple, anyone who can post anything is not reliable for anything. Similarly, Wikipedia articles are not reliable sources. Alanscottwalker (talk) 14:21, 29 January 2017 (UTC)
Anyone can self-publish something, or create a website claiming that it's the official website too. It's not as though "websites" go through some special validation process that Facebook accounts do not. (In fact, arguably the terms of service of Facebook are more restrictive than those of typical webhost services.) By extension then, official websites are not reliable sources. But we routinely reference official websites. So this suggests a degree of paranoia in evaluating sources that seems rather atypical for Wikipedia. Could you please clarify? Sławomir Biały (talk) 15:25, 29 January 2017 (UTC)
It has nothing to do with being paranoia. It has to do with ridiculously poor sourcing. Alanscottwalker (talk) 16:02, 29 January 2017 (UTC)
You did not answer the question. You wrote that "anyone... can post anything", which is no more the case on Facebook than it is on any official business website. In fact, it is quite possibly less so, since impersonation is explicitly against Facebook's terms of service, which they do enforce proactively. Sławomir Biały (talk) 16:14, 29 January 2017 (UTC)
If you have a claim against a company website, that you contend is unreliable for something take it to WP:RSN, here is not the place to discuss it.Alanscottwalker (talk) 21:52, 29 January 2017 (UTC)
Another evasion. If this is not the place to discuss reliable sources, then your belief that "It's rather simple, anyone who can post anything is not reliable for anything" should be raised at RSN rather than here. Since you appear to want to shut down rather than actually engage in discussion about a remark that you made, I will withdraw unless you do respond constructively. Sławomir Biały (talk) 22:38, 29 January 2017 (UTC)

On the general question: Yes, if an organization says something on their official Facebook page, we treat that just like the same organization saying something on their official website. There are ways to determine whether an account is the official Facebook page. The simplest is to look at their official website and find the link to their Facebook page. (This is trivially done in the instant case.)

Whether an organization's own official communication is reliable for a fact depends upon the contents. Generally, they are considered reliable for things that they say about themselves (including the people directly associated with them), but not for things unrelated to their own organization (e.g., climate change is/isn't happening, the Queen of England is a lizard-like alien, people who disagree with us will go to Hell, etc.).

All of which adds up to: Yes, you can cite the Facebook page to say that the railway bought something. WhatamIdoing (talk) 18:01, 14 February 2017 (UTC)

Yes, most in this thread, including me, agreed with this. But I raised another issue, a WP:TRIVIA issue with "someone bought something". And the original poster gave IMO a satisfactory rationale somewhat above. Staszek Lem (talk) 18:31, 14 February 2017 (UTC)

Availability

In Wikipedia talk:Identifying reliable sources/Archive 51#Archive or availability requirement for text sources, User:DES proposed that archiving and availability be considered as a key component for verifiability. It seems to me that WP:V and/or WP:RS once required that sources be "accessible" to the general public (e.g., not a case of a manuscript being locked into a lawyer's filing cabinet until the author has died, or the only known copy of a book is sealed under glass in a museum, or an unrecorded radio broadcast, etc.). Did we lose that language at some point? WhatamIdoing (talk) 04:30, 21 February 2017 (UTC)

  • Pinging @DESiegel:. I think the source should be checkable by a Wikipedian other than the author before it can be usable.—S Marshall T/C 17:37, 21 February 2017 (UTC)
  • Yes... sources need to be available to the public. Not necessarily easily available... but available never the less. Blueboar (talk) 18:46, 21 February 2017 (UTC)

IMO nice thought but a slippery slope for wiki-lawyering. Probably would be better that this be handled by a "strength of sourcing" metric in discussions where material that is sincerely contested for non-wp:ver reasons needs stronger sourcing. North8000 (talk) 19:01, 21 February 2017 (UTC)

  • If the only instance ("copy") of a source is "sealed under glass in a museum" (etc.), how could a Wikipedia editor have any knowledge of the contents in the first place? If an editor says "I remember hearing that unrecorded radio broadcast as a child" s/he is engaging in journalism; which is to say, making a first report in the public record. Such extreme examples do not inform this discussion.
It should be kept in mind that "available to the public" is inherent in the concept of "publication". The recurring issue here involves the sufficiency of "public" ("general"? or narrowly Wikipedians?) and "availability". The latter seems to be primarily about "paywalls". That can be a problem of access, but that is what libraries are for. Sure, many people don't have ready access to a top-grade libary. But then, many people don't have access to the Internet, which is in many respects just a bigger library. Either way, there are always "walls" of some kind, and nothing is universally and freely accessible to everyone.
I think the bottom line here is: if we limit the basis of our knowledge to (e.g.) what is freely available on the Internet, our representation of human knowledge would be very limited, and likely skewed.
As a separate issue, where an editor claims some obscure and even doubtful source which no one else can find, then a demand could be made to show the source. But we should not open the door to excluding or deleting material where access is merely inconvenient or has some cost. ~ J. Johnson (JJ) (talk) 21:11, 22 February 2017 (UTC)

RFC, "List of symphony orchestras in the United States" list verifiability

RFC,

Premise: Wikipedia is not a reliable source, verifiability requires reliable sources, and mainspace list articles require verifiability.
Conclusion: List of symphony orchestras in the United States is not an exception, as blue links are not reliable sources.

  • See related discussions at:
  1. #Can there be an exception case to allow blue links as references on lists?
  2. Talk:List of symphony orchestras in the United States#WP:V
  3. WT:Verifiability/Archive_61#Can lists of notable people be removed under WP:BURDEN?
  4. Talk:Elizabethtown, KY#Notable people
  5. WT:RS#Wikilinks, especially in list sections
  • Alternate views:  Alternate opinions already posted include that Wikipedia articles can be presumed to have their own sources, and that this is an elementary list.  If so, what defines this list as "elementary", such that sources can be presumed in the blue links?  Unscintillating (talk) 16:00, 20 February 2017 (UTC)
  • Agree as proponent.  We have a simple concept now, that blue links are not references.  If there are going to be exceptions, they need to be well-defined.  The possibility of ill-defined exceptions opens the door to broad scale removal of citations for specious arguments such as that inline citations are linkspam or bloat.  Unscintillating (talk) 16:00, 20 February 2017 (UTC)
  • Agree: If it's WP:MINREF material (and here it is, since it has been explicitly challenged), inline citations to reliable sources are a prerequisite for keeping the material. Each Wikipedia article must be able to stand on its own. In the context of verifiability, this means that the references that verify article contents must be present in this, instead of some other, article. – Finnusertop (talkcontribs) 16:18, 20 February 2017 (UTC)
    • If and only if it's MINREF material, then you're right. Except that in this case, it's not MINREF material, because the OP has explicitly said that he does not CHALLENGE the verifiability of even a single item in the list (to quote him: "while there is an interpretation that my actions were or are a wp:challenge, what I want is not based on a previous challenge"). WhatamIdoing (talk) 01:58, 21 February 2017 (UTC)
  • Weak disagree Addressing the point that this is an "elementary" list, I think it qualifies in that one of the defining features of the topic, the orchestra, is what city/area is operates from, from which it is trivial to determine if it is United States or not, so I really think that while we should encourage the sourcing for all blue-links, by definition of this topic, any blue-linked article that meets our notability guidelines will, by default, show why the topic meets this list topic, since part of the notability for any orchestra is where it is based. If it requires any more examination of the sources (say , a list of 80+ member orchestras, which is not a defining feature of an orchestra's notability), that would definitely require sources for each. --MASEM (t) 16:25, 20 February 2017 (UTC)
    • @Masem: If I'm understanding, part of this viewpoint is that there is information in the title of the article?  If so, do Lexington Symphony and Lexington Philharmonic Orchestra require citations?  Unscintillating (talk) 16:37, 20 February 2017 (UTC)
      • No, the title is not the factor here, though 99% of the time for an orchestra, the title will indicate where the orchestra is from. It is the fact that, working on the presumption that the blue-linked articles meeting WP:N guidelines (which thus assures WP:V), that in defining the notability of an orchestra is going to including defining its home, even if that is flatly obvious from the name of the orchestra. It always may not be, hence why I do not consider the title the indicator, but the sources asserting notability in the linked article. --MASEM (t) 16:51, 20 February 2017 (UTC)
  • I disagree with how the OP has framed the question and think it's not only indicative of their confusion as to just what the disagreement is about but also an unfair rhetorical tactic. No one is claiming that a bluelink should be used as a source. Nor has anyone questioned that verifiability should be required. These are straw men. The dispute is instead entirely over whether it is helpful to include inline citations in a list of articles that do nothing more than verify that they are orchestras and their location. No one has disputed that these facts can be verified for these entries, so let's not get misled into thinking this is about verifiability.
    Part of the OP's confusion seems to be an interpretation of "verifiable" to mean "accompanied by an inline citation", which is not what "verifiability" means. The presence of a citation gives a reader directions on where information can be verified, but does not prove that the information actually is verifiable, neither does the lack of a citation prove that it is unverifiable. postdlf (talk) 17:12, 20 February 2017 (UTC)
    • Saying that verifiability does not mean verified by an inline citation is correct. WP:V mandates that all material on Wikipedia needs to be merely verifiable. But we are dealing with WP:MINREF, a special case of verifiability that calls for certain content to be verified by an inline citation. I do not know if OP has confused the two or not, but the policies are not confused. MINREF material needs to be verified by an inline citation. – Finnusertop (talkcontribs) 19:02, 20 February 2017 (UTC)
    • Regarding the current contention, my last post summarized my position on the article talk page, and User:Postdlf saw no need to respond.  I stated, diff, "...my view is that unsourced material requires sources, and if not sourced may be challenged, although there is no deadline for either.  There is no consensus that there are exceptions that allow blue links to be used as reliable sources.  Unscintillating (talk) 02:21, 18 February 2017 (UTC)Unscintillating (talk) 19:41, 20 February 2017 (UTC)
      • The actual policy says that you can't CHALLENGE material unless you believe that it is unverifiable, and you've said that you don't actually intend to CHALLENGE the material or have any reason to believe that it's unverifiable. WhatamIdoing (talk) 02:07, 21 February 2017 (UTC)
  • If you read the linked discussions, you will see that I have been fairly consistent in my "do we need sources in list articles" arguments. So... I repeat all my previous arguments. But, to summarize and relate what I have said previously to the specific list in question: We need at least one source for each entry in order to establish two facts: 1) that the entity listed is indeed considered a "symphony orchestra" (we can not go by the name, it could be no more than a "town band" with delusions of grandeur)... and 2) that it is indeed located in the US. Blueboar (talk) 17:46, 20 February 2017 (UTC)
    • addendum - in most cases, there will be a good reliable source at the article on the individual orchestra that will establish both facts... easy to to copy and paste it for the entry at the list article ... it takes about 1 minute to do this... a LOT less time than it does to argue about whether the source is needed or not. Adding sources always improves an article. Take responsibility. It is pointless to argue about who should improve an article. Simply do so yourself, and don't worry about whether the other guy is doing so. Blueboar (talk) 20:40, 20 February 2017 (UTC)
  • Agree-IF those statements are challenged IMO the Rosetta stone is to start by looking at "what are the statements?" In this case, the statements are: That it's a symphony orchestra, that it's in the USA, and that it's notable. The first two are "sky is blue" statements, I would think that having a vetted-for wp:notability Wikipedia article should be enough to prevent a challenge by any reasonable person on notable for this particular list. Unfortunately, Wikipedia suggests but does not explicitly require at least a sincere question about verifiability to consider it a challenge. So, if challenged, normal WP sourcing is required, not WP itself. North8000 (talk) 19:51, 20 February 2017 (UTC)
    • Actually, the policy does require a sincere question about verifiability. The challenger is told to "state your concern that state your concern that it may not be possible to find a published reliable source for the content, and therefore it may not be verifiable". We added that a while ago, in response to some concerns about POV pushing and the occasional "I'm going to CHALLENGE every uncited sentence from here to infinity" discussions" (which have completely stopped since then, AFAICT). WhatamIdoing (talk)
@WhatamIdoing: I so so so wish that you were right, but in practice, I think not. As an indicator, one need only look at the two Feb 18th edits and edit summaries on the wp:ver policy page.North8000 (talk) 19:28, 22 February 2017 (UTC)
  • Agree per Blueboar. Regards, TransporterMan (TALK) 18:32, 20 February 2017 (UTC)
  • Comment This list gives a good example of why such lists may need to have some in-line citations. I scanned the list and looked for a name that might not fit the bill. The first one I looked at had citations that probably covered it, but Urbana Pops Orchestra has only one citation (""Millikin grad develops pop orchestra in Urbana") and that does not support UPO's the entry in the list, so it is not unreasonable to require a citation for that one. Personally I think that a blue link may be sufficient, particularly if the necessary citations are easy to find in the linked article, however entries such as UPO clearly need citations. Like Blueboar says it is usually easier to find a citation than to argue about it, but also there is no need to pre-emptively supply a citation for an entry like the New York Philharmonic because it is "[un]likely to be challenged" by anyone who is reasonably well educated (sky is blue). -- PBS (talk) 23:42, 20 February 2017 (UTC)
    • I'm classicly trained and I have no reason to know the difference between the New York Philharmonic and the New York Symphony Orchestra.  Moreover, there is something even easier than copying and pasting citations from a blue link, which is to not remove the citation that already exists, diffUnscintillating (talk) 00:38, 21 February 2017 (UTC)
  • Disagree with framing and conclusion. Unscintillating has said that no CHALLENGE is or was intended. He is unable to (in compliance with CHALLENGE) "state your concern that it may not be possible to find a published reliable source for the content, and therefore it may not be verifiable", because nobody has any reason to believe that this uncited material is actually unverifiable.
    The actual locus of this dispute is quite different: One editor really, really wants to have WP:Inline citations on a long list of undisputed entries, and another editor does not want the list cluttered up with a source that says, effectively, "Yup, guess what, something named 'MyCity Symphony Orchestra' is actually a symphony orchestra!" IMO this is an interesting question that the community has never addressed in any detail: Given a choice between providing inline citations for undisputed, unchallenged, known-to-be-verifiable information – or not – does the person who wants to add citations always win over the one who thinks that they're pointless clutter, including for WP:BLUE information, statements in the lead, etc.? WhatamIdoing (talk) 02:07, 21 February 2017 (UTC)
  • Strong Disagree - (edit conflict) First, I have to say (again) that I object to this enthymeme approach to RfCs, which seem more of an obscurantist exercise in circumscription rather than seeking clarity. We're left to fill in not just the practical evidence, but the practical effect, apart from the abstraction. Regardless, there's an obvious flaw in the logic. The "premise" leads one to believe Wikipedia is being used as a source. Obviously a link to a Wikipedia article is not a reliable source. This is made explicit in multiple locations, and there is no argument to the contrary. But that's only the misleading surface question, as what it's actually asking is the question that comes up frequently, framed in much clearer terms: "If a list article contains links to Wikipedia articles, do they also need inline citations, and if they don't have inline citations, should they be removed?" The answer, time and time again, is a big fat gray area. If you have reason to challenge the inclusion of a particular item, then go for it and put the burden on whoever wants to include it. Just saying "I contest everything that doesn't have a source, even if there are sources in the respective articles", however, is disruptive and contrary to the spirit of various policies (though to be clear, I know OP is not doing that). I'd be happy to continue this line of discussion in a separate thread. — Rhododendrites talk \\ 02:20, 21 February 2017 (UTC)
  • Disagree with premise I do think there is a "simple" way but not in the way stated. For such list, if you don't in good faith think an orchestra belongs because it cannot be sourced, either tag or remove it, per a proper challenge. If, on the other hand, you just want the source from the linked article to be transported to the page, SOFIXIT and transport the source. I do think any editor who sources can't be reverted for adding a proper source, at least on the basis that sources are "clutter". -- Alanscottwalker (talk) 00:20, 22 February 2017 (UTC)
    • Well, you might think so, but you may want to take a look at the edits that took place on that list on 5 February 2017. Search for the phrase "reference bloat" in the page history. And... it's maybe not that clear-cut in every case. The community might not support "any editor who adds sources", if they seem to pointlessly clutter up the lead. Lists are additionally complicated by history, because we used to ban sources from lists (and we still do ban them from disambiguation pages). WhatamIdoing (talk) 17:49, 22 February 2017 (UTC)
      • So? I still think so. It's idiotic to revert such sources. Alanscottwalker (talk) 18:00, 22 February 2017 (UTC)
  • of course a WL is not an RS. Wikipedia itself is not a reliable source. Jytdog (talk) 20:31, 23 February 2017 (UTC)
    • Of course it is not. However this is not an answer to the malformed FRC. Staszek Lem (talk) 20:50, 23 February 2017 (UTC)
  • Malformed RFC which has essentially the following format:
    Premise: Beating your wife is very bad.
    Conclusion: Beating your wife with a trout is not good.
I guess the underlying question was whether list items must be footnoted. Answer: Yes, when challenged. Corollary: In certain cases by the virtue of the list nearly all items may be challenged or blue links present challenge for verifiability, therefore the whole list may be challenged. Certainly we are not going to run around all List of Nobel Prize winners, slap tags onto them and delete items if not referenced in, say, a month. 20:50, 23 February 2017 (UTC)
  • Obviously agree that wikilinks are not a reliable source, but I would add that editors should not mass remove list items simply because there is no source included, but are expected to make a good faith effort to WP:PRESERVE apparently non-contentious material with all the usual BLP caveats. TimothyJosephWood 20:40, 23 February 2017 (UTC)
  • This is a great example of how not to write a neutral Request for Comment. If you want to get rid of all unreferenced list items on Wikipedia, propose an RfC about getting rid of all unreferenced list items on Wikipedia. If your sole issue is with one article's lack of references, fix the damn thing yourself. BrightRoundCircle (talk) 21:15, 23 February 2017 (UTC)
    • But Unscintillating tried to add citations to this particular list. The inline citations were removed by another editor with the edit summary "reference bloat". The practical point behind this discussion is "Please force the other editor to let me add these citations to the list, even though nobody, including me, CHALLENGEs even a single word on the entire page." WhatamIdoing (talk) 18:36, 24 February 2017 (UTC)
      • @BrightRoundCircle, WhatamIdoing, and Unscintillating: And on this note, let me add (since I've been vocal in my criticism of this RfC) that "reference bloat" is not a good reason to remove the sole reference next to a list item. When material is challenged, a citation is required, but the opposite is not true (that if material is unlikely to be challenged, citations should be removed). Ideally, every item in every list includes a citation, whether or not it's necessary. I've made an edit to WP:SAL to this effect, as it seems obviously in the spirit of WP:V and WP:CITE, and would welcome comments on that talk page. — Rhododendrites talk \\ 00:44, 25 February 2017 (UTC)
  • comment: the RfC question is very clear. It seems that the OP had some other point, but they didn't ask about that, so this RfC cannot speak to that question at all. From that perspective, this does seem to be a waste of time. Jytdog (talk) 21:23, 23 February 2017 (UTC)
    • There is no question in this RFC. There is no item actionable upon its outcome. Its statement is nothing but a tautology, therefore it is very clear. Convoluted answers are only because people trying hard to read between the lines or in the OP's head, in hopes to find any sense worth discussing. Staszek Lem (talk) 21:33, 23 February 2017 (UTC)
      • Oops. There is a question: what defines this list as "elementary", such that sources can be presumed in the blue links?. However it is poorly visible because it is bulleted out. Yes it is very clear, but only one person seems to had actually seen it, because "Agree" and "Strongly disgree" are not valid answers to it. :-) Staszek Lem (talk) 21:48, 23 February 2017 (UTC)
I don't like LIST articles and try to avoid them. The key thing for a LIST article is that the subject of the list is notable and per WP:LISTN you definitely need to provide multiple RS with significant discussion to have the list exist at all. Once the list exists, then individual items in the list of course need sources (as does everything in WP) as specifically described at WP:Source list. A WL is not a source. IronGargoyle set off this nonsense by (validly and usefully) removing what were essentially spam links to various orchestra's own websites, but instead of solving this simply by providing a high quality sources for each item (which should be very easy and fast) the folks at that article have gotten into this silly abstract argument. Just provide high quality refs instead of SPS and the issue is solved. oy. Jytdog (talk) 22:29, 23 February 2017 (UTC)
  • Most of the ~1/3 which were "Agree's" had one or two conditions that were not in the initial proposal. One was implicitly that it was to support the statements which are made in the header of the list and thus claimed by presence on the list. For example, the sky-is-blue statement of "is-an-orchestra" or the bigger-reach statement of "is notable". The other condition was "if challenged" with the two interpretations of what is required to / constitutes a challenge.North8000 (talk) 20:16, 24 February 2017 (UTC)

RfC Close challenged

Seriously? Closing your own RfC after 3 days with, effectively, "yup looks like I was right"? I would protest, but the same non-outcome would just be implemented by another user using language that didn't validate this RfC. Personally, I would probably opt for something like "Closed with consensus that nobody actually think blue links are reliable sources, so if OP wishes to seek clarity on something, he/she is urged to ask a direct question rather than engage in wheel spinning and/or rhetorical gymnastics." — Rhododendrites talk \\ 05:33, 24 February 2017 (UTC)

The RfC was closed by the proposer and their close summary does not represent consensus. Rhododendrites's summary is far more accurate. Unscintillating's close statement is possibly an attempt at WP:GAMING by later bringing up this "consensus" which isn't. BrightRoundCircle (talk) 08:29, 24 February 2017 (UTC)


I agree, the close should be overturned. I'm asking an uninvolved administration to overturn the close above. After that, let's please let the RfC run its proper course and close it appropriately. – Finnusertop (talkcontribs)

Agree. No harm, no foul, but need an un-involved person rather than the proposer. There were many comments on HOW it was proposed, which can mean issues that others can see rather than the proposer. So even a 100% effort at objectivity may not be enough. The closer indicated openness to overclosing or reclosing.North8000 (talk) 18:14, 24 February 2017 (UTC)
  • Close reverted I am reverting the close of this RfC as it appears to have been premature and done by an WP:INVOLVED editor. I suggest that a formal WP:RFC be posted using neutral language. -Ad Orientem (talk) 18:16, 24 February 2017 (UTC)

Proposed close

The RFC includes some brief very broad general statements, and then a general statement of applying them to a particular situation at a particular article. The broadness of the statements in the RFC led to discussion on several different topics in the same thread. Various concerns were expressed about the formulation of the RFC. One narrower application of principle was expressed by a good fraction of the participants and I believe not disputed by any others. That is if there is a question of the verifiability of the statements made by mere inclusion of an item in a list (e.g. in this case that it is an orchestra or that it is notable) then a source for the challenged statement must be provided. While variations of what is required to constitute a "challenge" were implicitly or explicitly expressed, this narrower application of principle met all of them. This narrow area where consensus was reached does not translate into any consensus on the statements/questions posed in the RFC. Suggest that anyone wishing to pursue this further should initiate a new RFC that is narrower and more specific on the details. North8000 (talk) 22:05, 26 February 2017 (UTC)

Reference bloat

"Given a choice between providing inline citations for undisputed, unchallenged, known-to-be-verifiable information – or not – does the person who wants to add citations always win over the one who thinks that they're pointless clutter"? (user:WhatamIdoing)

One person's trivial obvious knowledge is often not that of another person's, particularly as those who are native English speakers are from more than one country. Try playing trivial pursuits with natives in another country when they are playing with a game tailored to their country. Just one example who won the War of 1812?

The current wording in the section "Responsibility for providing citations" in this policy has developed to include a simple rule so that edit-wars over the inclusion of text must/ought to end with the include support that in-line citations to a reliable source if they are requested, has greatly reduced the drama that surrounds such disputes and usually ends up with the strengthening of the content of pages where inline cations are demanded.

For those of you who are not familiar with the reason for where burden lies, here a little bit of Wikipedia history. Back around 2005 it became a sport for news publications to pick up on academic concerns that Wikipedia had articles were inaccurate. It did a lot of damage to Wikipedia's reputation. To help guard against this there was a drive for quality over quantity. See:

There is a slow revert war going on at List of symphony orchestras in the United States of which the RfC in the section above called RFC, "List of symphony orchestras in the United States" list verifiability is/was a symptom.

Taking one "diff" mentioned in that RfC. It is example of something not explicitly covered in this policy. A case where an inline citation to a single source is removed with the comment "reference bloat".

In the past I have seen cases where there is information fully supported by more than one source where one or more inline citations are removed (or bundled into one inline citation) to avoid "the visual clutter of multiple clickable footnotes inside a sentence or paragraph", but never the wholesale removal of single inline citations that support a fact that is not common knowledge. I think that this issue needs to be discussed and if there is a consensus for it that explicit wording is either added to this policy or to WP:PRESERVE, as it seems to me to be to be the mirror image of WP:BURDEN.

Usually it the issue would probably be covered by WP:PRESERVE, but it is not if the editor who removes the inline citations honestly believes that they would not be included in the final article then it is not.

In some cases like this, I have seen a list supported by an initial inline citation that says something like "All the following items in this list are supported by Reference XYZ, unless noted otherwise". which I think is an expectable alternative if a lot of the items in a list are covered by one source.

As I said above in the RfC I do not think it is necessary to pre-emptively adding inline citations to facts that I think are well known, but I do not approve of deleting references just because they are considered clutter by one or two editors (it ought to take an article talk page consensus, before the revert of a deletion of citation to a reliable source is made -- this should include get out of Jail arguments like "this is sky is blue information", because if it is then it should be easy to get a consensus on the talk page that it is sky is blue information). This type of consensus is usually found for the removal of a {{citation needed}} on sky is blue information, if there is a dispute about whether it is sky is blue information (under the rules of BURDEN), so I think it reasonable that such a consensus should be needed to remove a <ref></ref> (ref-tag pair) on what someone considers to be sky is blue type information.

WP:BURDEN has the effect that User:Blueboar described in the RfC "- in most cases, there will be a good reliable source at the article on the individual orchestra that will establish both facts... easy to to copy and paste it for the entry at the list article ... it takes about 1 minute to do this... a LOT less time than it does to argue about whether the source is needed or not". But User:Blueboar what are you thoughts on the removal of reliable sources as highlighted in the diff given above? Also user:Rhododendrites you said above that you would be happy to discuss this further. So what are your thoughts on the removal of the citations as given in the same diff?

"Lists are additionally complicated by history, because we used to ban sources from lists" user:WhatamIdoing I do not recall when this was so. Can you link to the history of a policy page that said that?

Ping user:Unscintillating and user:IronGargoyle , but please do not rehash the details of your dispute here -- try to keep any arguments you present as abstract as possible.

Some guidance (but not policy), Wikipedia:Citation overkill; Wikipedia:You don't need to cite that the sky is blue and WP:LISTVERIFY

-- PBS (talk) 15:52, 24 February 2017 (UTC)

"Pointless clutter" is non-policy based objection to a source. But it's rather too much to have a discussion, here, about one article. Here is how this should have gone and can still go. Add source, revert source for some reason, discuss on the talk page, and reach an informal consensus or move to formal consensus (either RFC or RSN). The close of an RfC or RSN will presumably not credit non policy based objection. Now, if the same thing repeats itself over and over at different lists, you really do have much more of a case that there needs to be a policy (But the possibilities are 1) the same thing won't happen at most, or 2) if it very occasionally does, one can point to the last rfc. -- Alanscottwalker (talk) 16:26, 24 February 2017 (UTC)
  1. List of symphony orchestras in the United States doesn't need a thousand references, any and all information in that article can be sourced to a handful of sources.
  2. As for the general case, it's easier checking one good source that verifies hundreds of details than checking hundreds of sources, one for each detail. "Reference bloat" can easily be avoided by using few high-quality sources instead of many low-quality sources. BrightRoundCircle (talk) 16:27, 24 February 2017 (UTC)
That's all fine but it's not a reason not to have the discussion, at the talk page: 1) Add a source 2) someone reverts the source 3) come to agreement to use that multiple source -- end of . . . Alanscottwalker (talk) 16:31, 24 February 2017 (UTC)
Step 3 didn't seem to be working, as of the last time I looked at the talk page (~last week).
PBS, I found that rule in an early version of (if memory serves) one of the MOS pages on lists. It's been a long time since I saw it, so I don't remember the page, and I don't have time to look for it now. WhatamIdoing (talk) 18:41, 24 February 2017 (UTC)
That's when you go to RfC, on the talk page, or RSN. Alanscottwalker (talk) 20:39, 24 February 2017 (UTC)

Can I recommend [at least part of] this conversation move to WT:SAL? The idea that it's ever a good idea for an item on a list to have zero rather than nonzero citations of seems contrary to the spirit of WP:V and WP:CITE, but the language at WP:SAL along the lines of "generally expected that obviously appropriate material will not be supported by any type of reference" could easily be construed as "should not have a citation". I've changed it to say "... is not required to have an inline citation", which seems closer to typical best practices for citing sources/verifiability. — Rhododendrites talk \\ 00:50, 25 February 2017 (UTC)

I do not think that a guideline is the place to discuss this as this applies to all main space content, and ought to be addressed either in this policy or possibly in WP:EDIT. -- PBS (talk) 12:17, 25 February 2017 (UTC)

@Alanscottwalker, if someone adds text and it is deleted because there is no source, this policy is clear that the WP:BURDEN is on the person including text without sources to justify it. If there is an RfC and no consensus is reached then the default is not to include the text. It seems to me that if a person follows the advice here "If you think the material is verifiable, you are encouraged to provide an inline citation yourself before considering whether to remove or tag it." and provides a source then if someone removes the inline citation, then I think we need a follow up to that if someone removes the inline citation, not because it is not an unreliable source, but the footnote is in their opinion clutterb (or some other non policy reason). The burden to removing the source ought lie on the person removing it and if there is no consensus at an RfC the references should be retained. This I think is closer to the broader meaning of this policy, but it needs to be stated explicitly. -- PBS (talk) 12:31, 25 February 2017 (UTC)

I'm not sure that it really needs to be stated explicitly. This is the first dispute that I've encountered along these lines. Usually, re-writing a policy to handle a one-time dispute is WP:CREEPy. (If anyone knows about previous disputes along these lines, then I might well change my mind.) WhatamIdoing (talk) 05:27, 2 March 2017 (UTC)

List verifiabilty: "Year" and "Month Day" articles

The above discussions about list verifiability made my sight a bit more acute, and my eye caught a HUGE problem, which makes the issue of symphony orchestra almost trivial.

1948:

  • January 5 – Warner Brothers shows the first color newsreel (Tournament of Roses Parade and the Rose Bowl).
  • January 17 – A truce is declared between nationalist Indonesian and Dutch troops in Java.

Opinions? Staszek Lem (talk) 18:57, 24 February 2017 (UTC)

BTW, despite an unusually high share of IP editors in these articles, the level of vandalism is surprisingly low. Why's that? Staszek Lem (talk) 19:00, 24 February 2017 (UTC)

As with any content on Wikipedia, WP:MINREF material requires inline citations. Whether such entries are MINREF or not, varies. – Finnusertop (talkcontribs) 08:57, 2 March 2017 (UTC)

Amount of sources in Wikipedia compared to paper encyclopedias

Many Wikipedia articles have a bloated amount of separate sources as inline citations. Many paper encyclopedias use sources as bibliographies and seldom do inline citations. Sure, Wikipedia is not a paper encyclopedia; at least we have an option to print articles out. However, how do we compare ourselves Wikipedians to those doing research for traditional (paper) encyclopedias? --George Ho (talk) 18:32, 7 March 2017 (UTC)

Given the nature of the internet (or specifically, many internet users who choose to edit Wikipedia) I don't see any issues with holding WP to a higher standard for referencing than a paper encyclopedia. DonIago (talk) 19:51, 7 March 2017 (UTC)
True, each encyclopedia has standards and qualities on its own and may be individually high, though they may differ from each other. I'm not suggesting that referencing be standardized, and I'm not trying to oppose or counter your viewpoint. Just curious; that's all. And you're right; the nature of internet is increasingly complex... too complex to describe concisely. However, I looked over some articles on paper encyclopedia (and some articles at Encyclopedia Britannica online) and was marveled by the writing, organizing and research. Just curious, how would high amount of referencing affect readership and editing? Low amount of referencing? 1 − 2 + 3 − 4 + ⋯, a Featured Article, doesn't use much inline citations. However, many biographies of high-profile persons have extensive inline referencing. --George Ho (talk) 20:20, 7 March 2017 (UTC)
IMO completely different situations. Encyclopedias are generally written by experts. Wikipedia technically/ostensibly forbids writing from expertise. The latter is inherently going to require more referencing.North8000 (talk) 21:00, 7 March 2017 (UTC)
Kinda off-topic question: while Wikipedia does not allow "original research", how much OR has been inserted in print encyclopedias? George Ho (talk) 22:05, 7 March 2017 (UTC); edited. 22:17, 7 March 2017 (UTC)
Hmm... Seems already answered in Wikipedia:Expert editors#General? George Ho (talk) 22:08, 7 March 2017 (UTC)
Well, by the Wikipedia definition of OR, everything that is written from the writer's expertise is WP:OR; if you apply that externally, that would include the vast majority of material in other encyclopedias. But implicitly WP:OR is defined only for material in Wikipedia. North8000 (talk) 23:25, 7 March 2017 (UTC)
To come back to "amount of sources" (?): perhaps you mean how many sources need to be cited for a single point or fact? I've seen statements to the effect of "only one", which would be minimally sufficient. But where a point might be contested, though it has broad support, why not have multiple sources? In some cases the most solid, most apt source might be behind a paywall, or otherwise difficult to access, or even in technical language in an expert contest. So why not have multiple sources? In some cases there is an authoritative source (often a primary source), and having a secondary source that provides the wording an editor might use to describe the authoritative source seems very useful. Given all the foregoing, I think the question to consider is: what is a "bloated amount" of sources? Are there any examples worth looking at? ~ J. Johnson (JJ) (talk) 00:54, 8 March 2017 (UTC)
Some examples I might come up with are Madonna (entertainer) and Lindsay Lohan, which has a few hundred inline references. Syrian Civil War has nearly 1,000 inline references. I meant this "bloated". George Ho (talk) 02:35, 8 March 2017 (UTC)
If you have a good idea how to "compress" them, you are extremely very welcome. I don't know about Madonna, but Syrian CW does need every single word referenced, or you disagree? Staszek Lem (talk)
I do agree the texts like this look ugly, peppered with footnotes. One may come up with various ways of "sweeping them under the carpet". But unfortunately WMF is busy with various superambitious projects. Staszek Lem (talk) 03:21, 8 March 2017 (UTC)
Concur. The underlying problem is that WP's core policies necessarily require many references for complex or controversial subjects. For example, the Lawyer article (most of which I wrote) has to be so complex and so heavily referenced because too many countries have refused to adopt the simpler and more practical U.S. model of having a single unified legal services provider capable of practicing independently, so the article necessarily needs to be very long (and heavily sourced) to describe the sheer amount of crazy complexity around the world. --Coolcaesar (talk) 18:25, 8 March 2017 (UTC)
Apples and oranges. Wikipedia is not paper. This not the only thing WP differs from EB. Not even the most significant one. Staszek Lem (talk) 01:12, 8 March 2017 (UTC)
Syrian Civil War seems a useful example, with a number of aspects worth closer examination. E.g., that it has "nearly 1,000 inline references" – ah, what does "inline references" mean? The nearly a thousand sources that document the content? Or the "peppering" of the bracketed links (like: [515][516][517][518][519]) to the notes that contain the actual citations of those sources? (And before any says I'm making this complicated: no, I am not making this complicated, I'm trying to shine a light on certain complications that we keep tripping over.)
I think we can all agree that contentious points require at least one source. This is usually stuffed into a note (i.e., between <ref>...</ref> tags), which leaves the note-link "in-line". In this article there are often two or more such note-links concatenated. If all those adjacent notes were merged into single notes, the "peppering" of note-links would be reduced by about half.
The problem in doing such merges is that the standard way most editors do "in-line citations" – stuffing a source's full citation into the in-line <ref>...</ref> tags, or into a {{sfn}}, which amounts to the same thing – is not amenable to this. It basically ties each note to a specific source, preventing merging or mixing. The only way of reducing the note-links (the bracketed numbers) is to eliminate sources, which I think is a very bad way of solving what appears to be mainly an aesthetic issue.
However, there is a solution, a way of "compressing" these links, though many editors are set against it. That is the use of short cites, as implemented with {{harv}} templates. The nature of the sources here makes it challenging, and there would be a bit of work in converting, but it can be done. I think the issue comes down to: is the ugliness inherent in the extant citation practice reach the level where folks are willing to convert to a less intrusive practice? ~ J. Johnson (JJ) (talk) 21:36, 8 March 2017 (UTC)
Short citations don't reduce visual clutter, they add visual clutter—where once you had a link directly to the citation, now you have a link to the short-form citation that links to the full citation; extra work with no gain. The very simplest way to address this is with a CSS entry like .inline_footnotes {display: none}. The harder but better way to address this is to use fewer but higher-quality sources. BrightRoundCircle (talk) 21:52, 8 March 2017 (UTC)
Your assertion is nonsensical. To the extent that "visual clutter" refers to the bracketed note-links in the text, going from something like this [515][516][517][518][519] to this [515] is an undeniable reduction of "visual clutter". Whether use of short cites is more work for you depends on your work habits, which I can't speak for.
Simply using fewer sources is not a good answer, as an editor is presumably already using the best sources available. Where editors prefer to bundle all of their citations at the end of the sentence, it is highly unlikely that there will be single sources that cover all of the points in the sentence. Where such sources can be found it is unlikely they will be of the best quality. ~ J. Johnson (JJ) (talk) 06:28, 9 March 2017 (UTC)
"Bundling", in a different sense where one reference marker takes you to a "bundle" of references, is another way of reducing clumps of markers. Its merits and demerits have been discussed before, most recently here: Noyster (talk), 09:35, 9 March 2017 (UTC)
Short references do nothing to reduce the number of tags. They merely defer the page numbers to the reference section. What you demonstrated (reducing five citations to one) has nothing to do with Harvard references. Packing many references into one tag is "bundling" as Noyster says. BrightRoundCircle (talk) 14:11, 9 March 2017 (UTC)
"Bundling" is an ambiguous term applied to several different practices, some of which invoke strong feelings. Unfortunately, there is no general awareness of these differences, so lacking any definition of the term or clarification of what is meant (and even with a definition) some of you will react strongly on first mention of the term. Regarding some of these practices I fully agree. (E.g.: simply wrapping existing references in a {refn} template is wrong.) So please note: though the model I have described here can be considered a form of "bundling", please don't judge it by the faults of other forms which it is not.
BRC: how is it that you consider going from (say) this [515][516][517][518][519] to this [515] as "more" visual clutter?
It seems you do not understand what I am talking about. In particular, your term "Harvard references" is ambiguous. Perhaps you mean the "Harvard referencing" style, a form of parenthetical referencing, which uses a form of short cites in parentheses in the text. If that is the case, then I agree with you: what I demonstrated "has nothing to do with Harvard references", a form of referencing I neither use nor advocate, and have not even mentioned in this discussion.
I suspect you mistake the use of {{harv}} templates as implying the use of Harvard (or parenthetical) referencing. That is utterly incorrect. {{Harv}} templates create short cites, which can be used in various ways.
What I have suggested is replacing stuff like this:
  • <ref> <full citation> </ref> <ref> <full citation> </ref> <ref> <full citation> </ref>
which creates note-links like this: [1][2][3]
with short cites bundled in a single note (full citations collected elsewhere), like this:
  • <ref> <short cite>; <short cite>; <short cite> </ref>
which creates note-links like this: [1]
I am NOT suggesting mere substitution of full citations with a short cites, like this:
  • <ref> <short cite> </ref> <ref> <short cite> </ref> <ref> <short cite> </ref>
which makes no difference in the number of notes and note-links.
Is that clear on how the visual clutter of note-links can be reduced? ~ J. Johnson (JJ) (talk) 23:31, 9 March 2017 (UTC)
You completely misrepresented what I said. Read through it again. BrightRoundCircle (talk) 16:06, 18 March 2017 (UTC)

There's always talk about trying to make Wikipedia easier to edit. And the misguided efforts have been on taking the parts that are already relatively simple and making them simpler. Now y'all are talking about taking something that is essential and already complex and making it even more complex. And to solve a problem which generally doesn't exist, and which, in the few cases where it it does exist is there because of POV wars, which rage on due to other missing pieces of Wikipedia policies. North8000 (talk) 13:25, 9 March 2017 (UTC)

I agree with North8000 (and +1 point for using y'all). Regards, TransporterMan (TALK) 22:23, 9 March 2017 (UTC)
Thanks. I think that "y'all" fills one of the three most important missing words in the English language. A version of "you" that doesn't sound singular.North8000 (talk) 12:27, 10 March 2017 (UTC)
It is not clear to me what parts you consider "relatively simple" and what parts "complex", but that could be a whole other discussion. As the "bloat" Ho complains of, I would like to determine whether it is actually about the number of sources, or just the note-links that appear in the text. ~ J. Johnson (JJ) (talk) 22:35, 11 March 2017 (UTC)
When I said "bloated", I meant tons of bracketed links (or note-links), or "footnotes" (whatever you call them). Paper/Print encyclopedias seldom or occasionally do footnotes, don't they? George Ho (talk) 01:00, 13 March 2017 (UTC)
So we really are talking about the bracketed note-links (which link to the notes), not the sources themselves, and it's more an issue of consecutive note-links cluttering the text. As I just showed (above), multiple notes can be merged into single notes ("bundling", in some, but not all, senses of that term), resulting in just a single note-link at (say) the end of each sentence. Various objections can be raised on how to do that, but I think that comes down to just how referencing is done.
The more specialized encyclopedias often do footnotes, though not always to the extent of Wikipedia. The general, mass-audience encyclopedias seem to take the view "trust us, we know what we are doing (and get paid for it!)", and that notes would lose readership. Which is, of course, a reason for us to make the notes and note-links minimally obtrusive. ~ J. Johnson (JJ) (talk) 20:42, 15 March 2017 (UTC)
I would think that we should try to have editors seek to avoid having 4+ more cite refs in series, either by better attaching cites to the specific data/phrase they support and/or using grouped cites. But that becomes more a MOS issue, and only a bit about verification. In nearly every case I've seen of 4+ more cites, there's nearly always a way of rephrasing, rewording, or reformating information to avoid the 4+ cites in a row while dropping any cites and while keeping to WP:V. --MASEM (t) 22:29, 15 March 2017 (UTC)
Aaargh!!! I have only the vaguest idea what you might mean by "cite refs" and "cites", and your usage appears to be inconsistent. When you say "4+ more cite refs in series", do you mean something like this: [1][2][3][4][5]? (Those are the links to the notes, wherein we cite sources.) In that case I am with you (aside from exceedingly poor terminology), and I have already shown how to reduce such series of note-links to just one. Which doesn't require any "rephrasing, rewording, or reformating [of] information". If you mean something else: would you point us to an example or two? ~ J. Johnson (JJ) (talk) 23:02, 15 March 2017 (UTC)
Oh... I should have said number of "sources". When I said "brackets", I thought brackets and sources are interchangeable. --George Ho (talk) 19:15, 16 March 2017 (UTC)
Even I thought footnotes and sources or footnotes and bracketed links are interchangeable. I apologize for misunderstanding. The header and OP I made originally is about a number of sources in print encyclopedias vs. in Wikipedia. George Ho (talk) 19:19, 16 March 2017 (UTC)
You did say "sources" in both the section title "{{[tq|Amount of sources...}}") and your initial statement, though it was unclear just what you were referring to. Others have taken your comments as referring to the note-links, the superscripted and bracketed numbers that link to the notes that contain the citations that identify the sources. (Anyone who does not understand these distinctions should ask, as otherwise we lack a shared concept of what we are talking about.)
The brackets, and those bracketed numbers, are most certainly NOT the same as sources, not even as citations to those sources; these terms are not interchangeable. If you are concerned about series of "4+ more" bracketed numbers – that is, note-links — that is one discussion, and I have shown how such series or sequences can be reduced to a single instance. On the other hand, if your concern is about the number of sources – that is, the works from which an article's content is derived — that is a different discussion. You need to decide which topic you want discuss, and then clearly specify it. Discussion is futile if we all have different ideas of what we are talking about. ~ J. Johnson (JJ) (talk) 20:07, 16 March 2017 (UTC)

Reference inclusion

We have court documents that appoints my husband the title of Baron. How do I reference it within your guidelines Kazmac1312 (talk) 23:39, 19 March 2017 (UTC)

@Kazmac1312: You don't, per WP:SPS. Chris Troutman (talk) 17:41, 20 March 2017 (UTC)

Restart: do Wikipedia articles use too many sources?

My apologies to George, and I have renamed this to better capture what I believe he wants to discuss, and to make a cleaner break from the previous discussion. ~ J. Johnson (JJ) (talk) 23:03, 17 March 2017 (UTC)

My initial discussion is about the formatting and layout in print encyclopedias vs Wikipedia. However, my wording led to confusion, so I must apologize again. To rephrase my OP, I'll try this: we Wikipedians have been inserting so many sources to verify what we deem as challenging. The result has been hundreds or thousands of sources used. Why can't we Wikipedians be more like authors (i.e. academics) of the articles from print encyclopedia, like Encyclopedia Britannica (before it became online-only) and World Book Encyclopedia?

Although the discussion got convoluted, I got answers I needed the most. I'll summarize what I got: Historically, paper/print encyclopedias (whatever you call them) has been written by academics, as said by North8000. Number of sources have been at least one or 20. The articles from those encyclopedias in traditional (paper) format either rarely use or occasionally use sources as footnotes for referencing. Wikipedia, on the other hand, has a bunch of articles using a lot of sources. Some articles use 200 or 300 sources to verify info considered "challenging" to many. Even some others, like "Syrian Civil War", use nearly 1,000. The nature of internet makes Wikipedia unique and different from print encyclopedias. However, our common practices on verifying almost every info may have also alienated most academic experts who would have contributed the articles as much as we have done in Wikipedia. Those practices also made readership and editing less desirable and less attractive.

In conclusion, if we were like academics, then our articles would have used fewer (but reliable) sources than average. Also, there would be less footnotes, like how traditional encyclopedias do. However, due to Wikipedia's popularity, the nature of internet, and slow decline of print encyclopedias, the possibility to use less and less sources may be less likely. Maybe we can recruit more academics to turn to Wikipedia, but there are alternatives that would attract more academics without trying to recruit them. I hope I got this right this time. --George Ho (talk) 20:48, 16 March 2017 (UTC)

I think your real issue is with how WP's core policies evolved over time to respond to the challenges of a crowd-sourced encyclopedia that is a living electronic document and not a paper encyclopedia that sits for five years on a shelf in a public library. Articles in traditional paper/print encyclopedias can take it easy on the sources because they are usually written by one or two expert academic writers who can be trusted to write sentences where (1) if necessary, every assertion in that sentence can be traced to one of the sources cited in the article bibliography, or (2) the assertions are generally accepted as true among experts in that field and thus need no detailed sourcing. And if anyone scribbles nonsense by hand over those assertions, it's quite obvious to a user of a paper encyclopedia when the document has been vandalized.
It was found early on with the Wikipedia project that if Wikipedia accepts assertions too readily without any sources, it's way too easy for anonymous vandals acting in bad faith to inject flagrantly untrue nonsense into articles, and that can cause lots of trouble for innocent people when that gibberish consists of defamatory statements about living persons. That's why we have policies like WP:V, WP:NOR, and WP:NPOV --- the notions that assertions should be phrased as neutrally as possible and always subject to challenge unless they are verifiable restatements of the content of published reliable sources. Citizendium tried to create an encyclopedia through a more traditional expert-oriented writing process, but it has never been very popular. But I understand your frustration with how WP's footnote-obsessed writing style is also driving away many newcomers and experts who could otherwise make useful contributions to the encyclopedia. --Coolcaesar (talk) 23:24, 17 March 2017 (UTC)
Thou dost protest too much, methinks. In areas where Citizendium was able to attract scholars, their articles are far superior. And there is clear evidence of esteemed WP editors, some of our best, departing for that encyclopedia, because their solid contributions here became a morass of effort in trying to maintain them against an onslaught of inexpert editing. See for instance, the user page of the late (now deceased) chemical engineer Milton Beychok, AIChE, and his many contributions here and at Citizendium. Would I assign his works at either place to students to read? Most assuredly at Citizendium. Here, only if I had time to review them carefully, and who can give this time when an alternative, confidently reliable reading is available? (And therein lies the defeat of purpose, of this body of knowledge, and its claim to be encyclopedic.) Read the Shakespeare, if you are unfamiliar with the opening expression (adapted here to be gender neutral) or if you do not understand the take home message of this Talk reply. Leprof 73.210.155.96 (talk) 07:03, 18 March 2017 (UTC)
Is there a preference setting that makes Wikipedia display without any references? Maybe that will be what appeals to some folks who don't want to read an article with all these hanging footnotes? Or any semblance of a tag? AngusWOOF (barksniff) 01:17, 18 March 2017 (UTC)
The veracity of scholarly material lies in the scholarly reputations of its authors, and/or in the strong scholarly reputation of the editorial apparatus of the publishing house. Examples of this are the Encyclopedia Britannica, whose major articles bear the name of their authors (for instance, the EB article on "Coordination Compound" bears the name of the world renown scholar Jack Halpern, Prof emeritus in Chemistry at The University of Chicago) and Springer Verlag book publications (for instance, in the publication of Name Reactions by Jie Jack Li, whose credibility is established by the long scholarly publishing history of Springer, and the fact that Li is a professor with a fine academic pedigree, indeed, having co-published with a Nobel laureate in the field). In the sciences, the very same goes for publications from the AAAS, the Nature Publishing Group, Cell Press, the professional societies that publish the Journal of Biological Chemistry, etc. Other specific nuances apply in the social sciences and the humanities, but the fundamental point remains the same, even as the particulars differ. And what was said of Britannica is true of any good encyclopedia, general or specific (the scholarly credentials of the authors or publishers establish the credibility); see, for instance, articles at the Stanford Encyclopedia of Philosophy, or even the more local, historical Encyclopedia of Alabama.
Critically, at Wikipedia we have none of this going for us. All we have is that we cite sources, and maintain high standards of doing so. Articles that do not are plagiarism or WP:OR and have no credibility outside of Wikipedia itself. (It will eventually be the case, through expanding, critical scholarly efforts looking at the work of WP itself, that our own grades that we give articles will become public knowledge, and that external systems of grading will appear. When this happens, you will understand how poorly the academic world perceives the information given here—some few areas, such as Wikiproject Medicine, excepted—and you will, one hopes, stop arguing to dilute the little credibility it has, from adherence to its sourcing policies.)
But if these citation policies go by the wayside, then there is no hope of reliability, no basis for trusting us, and no remaining claim for our credibility whatsoever—and the policies are indeed increasingly sidelined, as individual editors ignore them, because we cannot hope to keep up with the corrections their sloppy work demands (and, I would suggest, increasingly sidelined as discussions like this take place and leave the waters more muddied than they were to start). "Just trust us" is already to frequently the unstated claim of articles (i.e., on any article with long tracts of unsourced material), and frankly, there is no a priori reason or strong justification for readers to trust what appears. Whether readers want the apparatus (superscripted footnotes) or not is immaterial—their existence is the only thing at present, absent the recognisable Jack Halpern and Jie Jack Li contributors or the esteemed staffs of Springer and the NPG, that we have going for us. Le Prof User:Leprof_7272 73.210.155.96 (talk) 06:44, 18 March 2017 (UTC)
Generally, agree. Thanks for laying that all out. Alanscottwalker (talk) 14:52, 18 March 2017 (UTC)
Yes, concur. A strong, and needed statement, which should be saved for future reference. (Thank you, Le Prof!) I think it is worth reiterating that reputable, trusted encyclopedias have trusted and expert edtiors, but we do not. Therefore we need explicit and visible standards and process. Anyone (newcomer or expert) that can't meet that standard has nothing useful to contribute. ~ J. Johnson (JJ) (talk) 22:14, 20 March 2017 (UTC)

It might be worthwhile to point out that if you're writing articles on largely non-controversial topics you often can do so with much less citations if you wish to do so. It won't be quite like academic writing but much closer to it.

Another thing to point out is, that it might be a bit doubtful whether sourcing requirments on their own really do drive away academics. Other reasons like often hostile content disputes, a rough discussion culture and the general nature of large scale collaboration (anybody might change your work) seem to be much more important to me. But all those troubles aside Wikipedia actually does have a lot of academic contributors which are responsible for much of its high quality content.--Kmhkmh (talk) 15:42, 18 March 2017 (UTC) --Kmhkmh (talk) 15:42, 18 March 2017 (UTC)

Sure, many people don't like to do in their spare-time what they do for a living -- the same is true of academics. Alanscottwalker (talk)
Yes. I find it quite doubtful that sourcing requirements drive away expert editors. Its more likely due to the rough treatment and lack of respect for expert opinion. ~ J. Johnson (JJ) (talk) 22:16, 20 March 2017 (UTC)
A little off-topic, but what is "expert opinion"? Is it more like WP:OR? --George Ho (talk) 07:46, 21 March 2017 (UTC)
It's the opinion of experts, those who have studied a subject well enough (typically with PhD degrees or similar) to be in the front ranks of those who know the most about a topic, and actually know what they are talking about. Not at all like WP:OR, which is where a WP editor (expert or not) comes up with material that goes beyond what any attributable source says, including the synthesis or conclusion derived from multiple sources. See WP:Expert retention#What is an expert editor?, and also WP:Randy in Boise for a contrast. Even experts are not allowed to present OR, but an expert is someone who is presumed to know both the subject and the field well enough to provide a viewpoint in line with critical mainstream thinking, as opposed to someone who claims *THE TRUTH!!!* based on what he has seen on tabloid tv or a few obscure reports. ~ J. Johnson (JJ) (talk) 22:45, 21 March 2017 (UTC)

Link to necessity and sufficiency

@Chris troutman: No reasonable person would think I made a material change to the policy itself in adding a link to this article there.--Jasper Deng (talk) 16:31, 20 March 2017 (UTC)

You were bold, I reverted, now we discuss. You should not feel empowered to make any change to a policy no matter how slight without discussion. Chris Troutman (talk) 16:39, 20 March 2017 (UTC)
@Chris troutman: So I have to get consensus for even a minor spelling error correction? Please be reasonable. The version as-is at this time doesn't even read differently.--Jasper Deng (talk) 16:40, 20 March 2017 (UTC)
We don't allow wikification within quotes for fear that it would alter the quote in an inaccurate way, even though there are many cases where wikification would explicate the matter for the reader. Just the same, I oppose you adding a link as it adds particular content for which you do not have consensus. Again, why would you think that policy pages are something you should edit on a whim? Chris Troutman (talk) 16:43, 20 March 2017 (UTC)
@Chris troutman: Did you really think I would do something like replacing "requires" with "does not require" in the lead? Of course not. This is nothing like that. That's nothing but a red herring here.
And as the reverter, the burden is on you to give a valid reason why this edit should not be made. Per Wikipedia:Silence and consensus I presumed I had consensus for a minor edit of this sort. And an edit that clearly does not change the logical content of the sentence nor its intended meanings and connotations falls in that category. Not allowing wikification in quotes is another red herring: a policy is not nearly as loosely written as someone's personal quotes, and in this case does not suffer the potential problem of changing the quoted speaker's intentions.
Many people are confused about something being necessary versus being sufficient. The only difference in intended meaning and wording in the policy's language is "require" instead of "necessary" and "reason for" instead of "sufficient". The meaning of the policy is completely captured by the article.--Jasper Deng (talk) 16:50, 20 March 2017 (UTC)
@Jasper Deng: Neither WP:SILENCE nor WP:BRD place any burden upon me except providing a reason other than IDONTLIKEIT, which I have. Per BRD, "if you voice dissent, failure to make your dissent heated and continuous does not constitute silence". I don't need to discuss this further. Just to make plain to you, your description of red herrings is a blatant dismissal of community practices. That you don't agree with them does not invalidate them. The article about "necessity and sufficiency" only says what it says now. That could change. Further, that you think someone is confused about the concept still does not give license for adding a link from this policy page. Per WP:NOTPART, "the content of these pages is controlled by community-wide consensus". You can't even manage to convince me, let alone everyone else. So, you're welcome to start an RfC, seek a third opinion, or just move on to editing somewhere else. Chris Troutman (talk) 17:39, 20 March 2017 (UTC)
@Chris troutman: I never challenged your invocation of BRD, or I would have had issue with your second revert. What I was challenging was your supposition that I must get consensus for each and every edit to project namespace pages, which is blatantly and clearly false.
And your admission that you had no reason for reverting besides that and Reyk's support below shows that the only effect of your reverts has been obstructive, even if unintentionally. It appears (even if you're not intending to) you are wikilawyering about NOTPART. If it were literally true, then even fixing a slight spelling error needs consensus? Clearly not.
If the article I linked to changes significantly in the future (which I very strongly doubt it would), and becomes inappropriate to link, then someone can remove it with that reason in the edit summary.--Jasper Deng (talk) 19:02, 20 March 2017 (UTC)
  • I don't see a problem with Jasper Deng's link, and I support leaving it in. Reyk YO! 18:12, 20 March 2017 (UTC)
I'd agree. This particular link is helpful in this particular policy. Chris is correct that significant changes to core policy pages such as wp:V do need consensus. It is not just any "project namespace page", as Jasper put it. However, this change is minor and constructive. Unless there's a clear problem identified with the change sometime soon, Chris should self-revert the reversion. LeadSongDog come howl! 19:13, 20 March 2017 (UTC)
Yes, thank you. My entire point is this: I don't need consensus for minor changes that clearly don't change the content of the policy, whereas I of course recognize that if I had completely reversed the meaning of the sentence or otherwise made a significant change to the logical content, of course I would first discuss it here.--Jasper Deng (talk) 19:16, 20 March 2017 (UTC)

There were several stages to the edit so it's not exactly clear what is being proposed, but this would apply to most of them. That sentence was just the result of a large amount of work on the talk page. Also, the linked page obscures the point that is relevant here rather than informing on it. North8000 (talk) 19:51, 20 March 2017 (UTC)

(edit conflict) @North8000: And how so? Not to force you to be a logician, but please name me one instance of something covered by this statement on Wikipedia where the original statement had different truth value than afterwards. I'm merely proposing to wikilink necessity and sufficiency, without the "alone" I had mistakenly inserted. Edit: I see you removed your assertion that it was a "major change" and so the burden is no longer on you to justify how.
As for informing a reader, believe me, plenty of people conflate requirement with eligibility (i.e. necessity with sufficiency). This is why we have to explictly say in WP:NOT that we aren't an indiscriminate collection of information.--Jasper Deng (talk) 19:53, 20 March 2017 (UTC)
I like your reasoning, but I think that the link confuses and weakens the (relatively new) sentence rather than enhancing it. North8000 (talk) 20:09, 20 March 2017 (UTC)
@North8000: My interpretation of the sentence is that just because a fact is verifiable doesn't mean it is automatically eligible for mention in Wikipedia. My belief that the link is a clarification rather than a confusion is founded on that interpretation. If that interpretation is wrong, then I don't guarantee my argument's validity. But under that interpretation, I don't think the sentence adequately conveys the difference, and that the article helps clear it up.--Jasper Deng (talk) 20:12, 20 March 2017 (UTC)
@Jasper Deng:There was recently a substantial discussion on this sentence. Of course I encourage reading the whole thing, but IMO the key points are incorporated below. The goal here is to establish the requirements imposed by wp:ver and also to end a mis-use which is in essence saying (in disputes) that meeting wp:ver is a force for inclusion. With respect to your post:
  1. In wp:ver we should avoid venturing into talking about sufficiency for inclusion. Suficiency, besides complying with zillions of other policies and guidelines, is also a highly variable standard. On a small one editor article for an uncontested item, that bar may be down near zero. At the other end of the spectrum you might need to get a strong consensus in an RFC to put it in.
  2. Even if "sufficiency" were germane here, the linked article confuses rather than informs. It starts from the general case where "necessity" is undetermined, whereas in wp:ver it is a given, a constant.
  3. The main problem is mis-use of wp:ver in disputes, it's not one of there being too low of a bar for inclusion. So IMHO talking "sufficiency" has the additional problem of derailing it from the main point of the sentence which is that meeting wp:ver is not an argument for inclusion.
Sincerely, North8000 (talk) 21:18, 20 March 2017 (UTC)
Jasper Deng: your edit did, in fact, did make a change to the language, even if it was small, and your lead comment that no reasonable person would think so is hardly helpful. Considering how much time and trouble the rest of us took to hash out the often subtle nuances, it was rude of you to just slide in and change things. And upon being reverted you really should have left it alone. Throwing it right back in demonstrates a poor attitude and lack of collegiality which rather dissuades me from addressing any questions and concerns you have. The tone of your comments (including the imputation of wikilawyering) definitely puts me off. I concur in full with what Chris has said and done, and likely with whatever he adds. North8000 also seems right on. ~ J. Johnson (JJ) (talk) 21:32, 20 March 2017 (UTC)
  • Strongly oppose to J.D.'s changes. Wikipiping is a quite abusive practice in many places in wikipedia I often run into. Whatever was wikipiped by J.D., is way far from being synonymous. Staszek Lem (talk) 01:57, 21 March 2017 (UTC)
  • Seems like an innocuous change to me and I don't see any reason to dispute it. However, it's drawn my attention to the awful wording someone's introduced into the lede. If we want our encyclopaedia to be a source of accurate and helpful information then verifiability certainly is a reason for inclusion, provided the information being verified is introduced in an appropriate place and given prominence in proportion to its importance. So that sentence is way wide of the mark. I'm also opposed to the silly edit to WP:ONUS, which changed the rather precise and well-thought-out word "information" into the much less specific "content". It really is the information that needs to be verifiable. If the underlying information is accurate, appropriate and proportional then we really don't want people invoking WP:V to change images or other files, or to dispute specific wording choices.—S Marshall T/C 18:20, 22 March 2017 (UTC)
    • I am afraid we are dealing with non-binary logic case here. Non-verifiability is a reason for exclusion 100% regardless anything else. Whereas verifiability may be an argument favoring inclusion, but it is not 100% enforcing inclusion (like you said yourselves, "provided..."). Re: So that sentence is way wide of the mark -- If you are talking about the prev version from this restorational edit, I also support it. Staszek Lem (talk) 19:38, 22 March 2017 (UTC)
    • Re: silly edit to WP:ONUS -- I agree that 'information' and 'content' are not exactly synonymous, with "content" being a form of the presentation of information. Therefore "verifiable content" is a rather hapless phrasing here: "verifiable content" would rather mean that the content faithfully represents information. Of course, this is an important issue, but it is one step removed from this policy, which requires that the information comes from a reliable source; we don't demand that "content comes from reliable source", which will be a snake pit/can of worms wide open, as S Marshall noted, unless it is an exact quotation of the source. Staszek Lem (talk) 19:50, 22 March 2017 (UTC)
      • I dug into history; and noticed that Sławomir Biały's partial restoration introduced the discrepancy, so I completed the revert. See:
        • (cur | prev) 12:59, 22 March 2017‎ Staszek Lem (talk | contribs)‎ . . (33,273 bytes) (-37)‎ . . (→/*‎Verifiability is not a reason for inclusion*/: the prev restoration was incomplete; this and prev are revert of [1]) (rollback: 1 edit | undo)
        • (cur | prev) 11:39, 22 March 2017‎ Sławomir Biały (talk | contribs)‎ . . (33,310 bytes) (+57)‎ . . (→‎top: restored previous wording) (undo | thank)
      At the same time the reverted edit (with summary "Adding a clarification per (more or less) the discussion at WT:Verifiability#Proposal") and the next one had the problem in the "more or less" part. The edit omitted an important word the editor was favoring themselves: "not a sufficient reason for" -- omitting "sufficient" created logical blunder just discussed. Staszek Lem (talk) 20:06, 22 March 2017 (UTC)
IMO it was not a logical blunder, and that including "sufficient" would have been a logical blunder. This was discussed a great deal and so I won't expand here unless someone wishes. Sincerely, North8000 (talk) 20:52, 22 March 2017 (UTC)

In this edit of 20 March Jasper Deng added to the lede a link to Necessity and sufficiency, and the word "alone". The latter altered the meaning of the text, but he subsequently removed it, so I think that should be considered an error in editing that was corrected, and therefore not in issue. The issue of this section is therefore the link, and I hope that any extension of the issue as to the broader wording will be reserved for the following discussion.

As to the link: I most certainly can appreciate 1) the distinction between "necessity" and "sufficiency", 2) that many people often overlook that distinction, and 3) the applicability of that to the text. HOWEVER, the article linked to is a rather technical, even overly erudite, treatment of the concept that does nothing to clarify or explain the text is is applied to, and most likely would only befuddle anyone clicking on it. Quite aside from whether any link might be helpful (let alone is necessary), this link is not only not helpful, it is dismally unhelpful. ~ J. Johnson (JJ) (talk) 07:02, 23 March 2017 (UTC)

Please clarify—authoritative responses only (admins and experienced editors)

Please comment here, clearly, as to the meaning of the bolded expression, "about themselves," appearing in the section WP:ABOUTSELF. It is being used to argue for the validity of Facebook and Twitter as sources for BLP information on a subject, when the Facebook pages and tweets contain self-referential (autobiographical, self-published) statements about the subject of an article. Please state here, and as necessary, clarify in the text of the relevant guidelines/policies, when it is, and is not suitable to use a tweet or FB posting from a person, to populate their BLP biographies at Wikipedia. Thank you in advance for replies. Cheers, Le Prof User:Leprof_7272 73.210.155.96 (talk) 06:52, 18 March 2017 (UTC)

As one of the editors who helped write WP:ABOUTSELF... I can answer with: not originally. We were thinking more about personal web pages and blogs. Thinking back, I am not sure that Facebook and Twitter even existed back then (and if they did, they were certainly not as ubiquitous as they are today). Of course that was a long time ago, and the situation may have changed. What would concern me is whether Facebook and Twitter posts are permanent enough for real verifiability. Blueboar (talk) 11:00, 18 March 2017 (UTC)
@Blueboar: And yet, B, these two social media venues, and others, are now explicitly mentioned, and so it has to be the consensus of admins and editors here that WP:ABOUTSELF applies to these. The question is how does it apply? Does it mean that anything a person says on their Twitter feed to Facebook page—"I am releasing a new album." or "All Republicans are fascists." or "I am on a date with Justin Bieber." or whatever—is valid content for Wikipedia? What does this bold phrase mean? When are self-referential statements in these social media acceptable as sources, and when not? Certainly guidelines and policies had their start before the rampant self-publication era, but they have evolved since. What is their meaning today? Cheers, Rsvp. Le Prof User:Leprof_7272 Leprof 7272 (talk) 18:22, 18 March 2017 (UTC)
I don't think the Twitter and Facebook posts need to be removed, given that they are considered to be extra sources incase if we don't find the original article!!!--AnimeDisneylover95 (talk) 12:09, 18 March 2017 (UTC)
In my view Facebook and Twitter posts should be used very carefully if used at all. People do not use social media to tell the truth about themselves. In general, social media profiles are selective, sanitised, and idealised, and they present a distorted picture. I would suggest that social media should only be used as a source about a person if there is a talk page consensus that they are appropriate for the particular fact being cited. If there's a dispute between good faith editors then I would recommend that they are left out.—S Marshall T/C 12:35, 18 March 2017 (UTC)
@S Marshall: Thank you for your input here. Basis for opinions in WP policies and guidelines? Cheers, Le Prof Leprof 7272 (talk) 18:22, 18 March 2017 (UTC)
At the moment you're looking at a conversation between editors, rather than an edict from on high. I'm seeking consensus for my view rather than necessarily declaring that it must be as I say. I think it's likely that North8000 and Blueboar will concur with me to a substantial extent, though. My position is that a claim about a living person based only on a social media post may be removed by any good faith editor and should not be replaced unless there is a consensus to put it back. This is based on the second paragraph of WP:BLP. The key phrase there is reliable, published source. Social media posts do not meet the definition of a reliable source in WP:RS because of the lack of fact-checking.—S Marshall T/C 20:12, 18 March 2017 (UTC)
Hold on... based upon what AnimeDisneylover9 said above, I suspect that this isn't about using Facebook and Twitter posts in general... but about using a specific post/posts in a specific article/articles. Could we please have the details?... a link to any talk page discussions would be helpful. Blueboar (talk) 14:13, 18 March 2017 (UTC)
Note, fellow editors: In mentioning "the Twitter and Facebook posts" [emphasis added], @AnimeDisneylover95: is referring to a brief three party discussion at the Talk of Cristina Vee, which is immaterial to the general question being asked here (and which belongs and should remain at that article Talk page). There is no editor war there—editor AngusWoof is busy editing at that article, and he and I have traded off doing so today. While that article was where the variant interpretation was raised—when Angus reverted some edits that were Tweet-only sourced content—there is no active discussion there to join. Persons interested in looking in at the other article, are welcome to do so, here, and here—but the matter I want to raise is a general one, and so is raised here.
I once again ask for material limited to interpretation of the guidelines relevant to social media within WP:VERIFY, and specifically to the meaning of WP:ABOUTSELF. Le Prof User:Leprof_7272 Leprof 7272 (talk) 18:22, 18 March 2017 (UTC)
The policy section that you are asking about is reasonably clear, and it specifically mentions Facebook and twitter, and so the wording is relevant to them, and it is in WP:VER which is not only a policy, but one of the core / most authoritative policies. Policies get interpreted for particular situations, the specifics of the situations are very relevant to the answer/interpretation, and people are ready to help do that. But you are saying that it is not about any particular cases even though you opened with "It is being used to argue for the validity of Facebook and Twitter as sources for BLP information on a subject," Given all of that, what's left? / exactly what are you expecting to get here? North8000 (talk) 19:32, 18 March 2017 (UTC)

The article has been overhauled to limit where the subject has tweeted her upcoming roles in shows and video games. Many of those are of the "I voice (character) in (video game)" types of announcements, and were either replaced by secondary sources, reviews or if primary, press releases (and yes, sometimes tweets by the directors) of cast. The question then becomes whether a profile that is posted for a convention is considered self-published, besides the ones that cry "come and see me at (convention)" or ones where they copied Wikipedia or a common bio from a website. Then there's the question of whether a writeup by a publisher of the subject (in this case the audio book company, the video game company, or the film company) is also considered a self-publish. After stripping most of those out, what's left now for her self-pubs are a set of tweets concerning her own birth month and day (a string of annual tweets thanking people for the birthday wishes, plus some tweets stating how her birthday is the next day, see talk page), birth place (2 tweets, one saying she's born in the US, and one where she responds "born and raised" to whether she was from LA), and current residence (tweets announcing her move from home town and actual arrival). I think those are good enough to not require tagging the entire article or the inlines. They are not extraordinary or contradictory claims. Her Twitter account is verified. Is there still a concern? AngusWOOF (barksniff) 20:52, 18 March 2017 (UTC)

@AngusWOOF: the only matter for discussion here is the interpretation of WP:ABOUTSELF meaning, in the broader context of WP:VERIFY (e.g., with its WP:SELFPUBLISH section) and the separate WP:INDEPENDENT article. I applaud your work at the CV article, and am not trying to discuss all issues of citations here. (Those are better discussed at that article's Talk page.) Obviously , the consensus that appears here will be relevant to your broader questions. Of what you mention, the aspects of all
I haven't looked into this particular article. As a general point of principle: yes, I'm still concerned. If the only source for a claim is someone's social media account, then I would seriously wonder that claim was important enough, and verifiable enough, to include on Wikipedia. Surely if someone's birthdate or business activities mattered, a secondary source would have picked it up. Yes, I know infoboxes have parameters for a lot of this information but we don't need to populate infoboxes and sometimes it can be good editorial judgement not to.—S Marshall T/C 14:24, 19 March 2017 (UTC)
I imagine all are back at the real work of life, and I appreciate if closing this discussion will entail a delay. North8000, this discussion has developed how I hoped, providing a relatively general discussion of how FB and Twitter should and should not be used. I do yet think that the following matter of interpretation of the wording of the WP:ABOUTSELF section should be discussed. I perceive that the content of the section is being interpreted two ways:
  1. either that FB pages and Tweets can be used for BLP, to generate content about article subjects, as those subjects self-disclose in those social media pages, about themselves,
  2. or that FB pages and Tweets can be used to substantiate only that a statement has been made regarding a subject, at these social media pages, and not much more.
An example of 1. is the case mentioned by Angus (that Cristina Vee tweets about her birthday and her changing place of residence, and so her tweets can be used to substantiate her DOB, and her place of residence as of the Tweet date). Applying the principle of 2. to the same case, we can only report something like "Cristina Vee has discussed her birthday in her tweets" (citing such a tweet), etc.
I hold to the second interpretation, because if the first interpretation holds, it contradicts WP:SELFPUBLISH in W:VERIFY, which prohibits/proscribes use of self-published content. So again, @Blueboar: @S Marshall: @North8000:, one last weighing in—how does one interpret this particular aspect of WP:VER? Are either of the two interpretations offered valid, or is there another? If you need a particular case to apply it to, look to the description by @AngusWOOF: of his editing of that Cristina Vee article—where he has taken it far in the direction of reliable sourcing, removing reams of material that is questionable in terms of sourcing or notability (including removal of many CV-authored tweets).
Bottom line, an issue still appears to remain, somewhat, for some involved here, over this WP:ABOUTSELF section, in relation to the WP:SELFPUBLISH and other related sections of WP:VERIFY. Can one not be accepted to state their birthday (i.e., is it not okay to cite tweets when people talk about themselves)? I think this flies in the face of WP:SELFPUBLISH, but also WP:INDEPENDENT, but I, personally, would allow a cited, self-published Tweet if it is in addition to a third-party independent source, as the Tweet then strengthens the third party report (and vice-versa), because I do not see this as violating WP:VER or WP:INDY. What say ye, on this? So, I look again, here, for greater clarity on this particular matter of interpretation (using the Cristina Vee article example, if necessary).
Also, @Blueboar: @S Marshall: @North8000:, @AngusWOOF:—are there other esteemed admins and editors we would like to invite to comment on this, so the matter is more clearly established? If and when it becomes a broader issue, we could reference this at or move this to a higher level discussion/noticeboard, but in the meantime, it would give us a discussion to point to where the nuances of the issue have been thoroughly discussed. Cheers, Le Prof 73.210.155.96 (talk) 14:51, 20 March 2017 (UTC)
I wanted to state the LeProf was also contesting tweet announcements by the show's directors and tweets from official website accounts about casting announcements, saying those weren't third-party independent enough either. Also, what about when the actor posts a correction about their birthday on social media? See Jason Statham's page, where he has corrected the month and day and that's garnered lots of news articles in reaction to it. BTW, contesting his birth year is another story since there are a bunch of media articles supporting either version, and Wikipedia's even been mentioned in articles about the debate, and it'd be a bit like contesting Barack Obama's place of birth. AngusWOOF (barksniff) 17:36, 20 March 2017 (UTC)
@AngusWOOF: I plead no contest to the above. I am simply trying to keep matters focused on the general concerns, which, if addressed, would speak to all these specific cases. @Blueboar: @S Marshall: @North8000: — again, the matter is not about this one article, but about arriving at a general understanding of what WP:VERIFY means when it discusses ABOUTSELF, and what it and related policies wish us to do about self-published social media in general. Regardless of whether twitter accounts are verified, they constitute a person talking about a matter in which they are involved, and have a vested interest. Hence they are autobiographical, and not independent/third-party. As I have said, I, personally, would allow a cited, self-published Tweet if it is in addition to a third-party independent source, as the Tweet then strengthens the third party report. But in summary, it would be helpful if earlier participating editors would skim the above, and give their final thoughts. Please, Blue, Marshal, North, others, help bring this discussion to a practical close. Le Prof Leprof 7272 (talk) 20:35, 23 March 2017 (UTC)

@AngusWOOF:, others: Discussion closed here, new section begun below, with a firm proposal on the table, with a view toward a Roberts Rules of Order approach, whereby the proposal may help us move to a consensus on how to view this issue. Thank you all. Le Prof Leprof 7272 (talk) 20:54, 23 March 2017 (UTC)

Another RfC discussion at Talk:Cold War II

I started another RfC discussion, Talk:Cold War II#RfC: "Novel risks and measures for preventing escalation" section. I invite you to comment. --George Ho (talk) 07:26, 5 March 2017 (UTC)

The discussion is relisted. --George Ho (talk) 07:00, 29 March 2017 (UTC)

Where verifiability is not an adequate criterion

In the previous section the objection to the disputed material was not one of verifiability (the claim made has an RS), and, as we have seen, reproducibility is not a suitable requirement. Still, I wonder if there might be a proper objection, which it might be useful to identify. I see three broad possibilities.

  • First is the determination of the mainstream view (our mandated point of view) and whether specific material (especially claims) is in accord with mainstream opinion. (Caveat: I usually work on scientific topics, where the relevant mainstream view is that of science, not the mass media.) This can be challenging, for reasons similar to those mentioned in regard to reproducibility. Some aspects of this are covered under WP:FRINGE, but that doesn't really cover novel results (often cried up in the mass media), which sometimes are truly break-throughs, but often simply not reproducible.
  • The second broad objection is a condition that I am sure we all recognize, but does not seem to have been taxonomized and given a name. This is the tendency of many editors adorn articles with "interesting" facts, factoids, distant connections, etc., which do not contribute to, and often distract from, a more considered presentation of the topic. I propose we call this "pack-rat ornamentation".
  • A third class concerns the proper weight to be given to subordinate details. This is related to, but should be distinguished from, WP:WEIGHT, which covers the proportioning of attention given to different points of view. The basic idea here is that given a certain sized article, with a breadth adequate to cover the topic, there is a limit to the depth of detail that can be covered, and that, generally, no subtopic or detail should be given more coverage than can be given to all of the subtopics at that level.

Comments? ~ J. Johnson (JJ) (talk) 23:07, 7 February 2017 (UTC)

Verifiability alone is not (and never was) a sufficient criteria for mandatory coverage just a necessary one. However the inclusion/coverage of content falls partially in the realm of editorial discretion, which we should not and imho cannot do away with. Personally I'm rather skeptical in creating more and more rules & details to guide that, this can easily turn into WP:CREEP.--Kmhkmh (talk) 02:10, 8 February 2017 (UTC)
1st: You are correct that "mainstream" or not is covered in WP:FRINGE / essay:WP:Mainstream / essay:WP:SCICON As for "cutting-edge"/"breakthroughs", IMO I got an adequate explanation in the section above. Staszek Lem (talk) 03:18, 8 February 2017 (UTC)
2nd: factoids & tangential are generally covered by WP:TRIVIA/ WP:UNDUE. Staszek Lem (talk) 03:18, 8 February 2017 (UTC)
3rd: Level of detail / subtopics is covered by WP:SUMMARY. Staszek Lem (talk) 03:18, 8 February 2017 (UTC)

It's common for editorial discretion to get hampered by a widely used myth. Maybe a statement like this which is a clearer summary would deflate the myth and let these things work themselves out better: "Verifiability is a requirement for inclusion, not a reason for inclusion". North8000 (talk) 04:30, 8 February 2017 (UTC)

Is this not already covered under WP:Verifiability#Verifiability does not guarantee inclusion? Blueboar (talk) 13:16, 8 February 2017 (UTC)
IMHO no, it is not. Structurally, "not a guarantee" is a very narrow and thus a very weak statement. It leaves it open that verifiability can be considered to be a force or argument for inclusion. Now, on to "in practice": Here's an example of how widely accepted this myth is. Let's say I put in some item which is sourced but a really bad idea or really bad for the article. (and let's say that no other policies or guidelines clearly rule out my insertion). Then John Doe undoes my insertion giving a editorial discretion type reason. Now is when the editorial discretion discussion should start. But let's say then I revert his removal and I say "Undo removal of sourced material", and say similar things in any further discussions. The ONLY argument I have made is essentially that verifiability is a force or reason for inclusion, so I have given no other basis for keeping it in. Yet this is widely considered to be a valid argument in Wikipedia despite per se having no basis in policies or guidelines. Hence me terming it a "widely accepted myth". North8000 (talk) 13:44, 8 February 2017 (UTC)
It's a prima facie valid reason for putting in an edit summary of a revert of a removal of content, in my view, but if the editor who wants to remove the content has what they believe is a good counter-argument, they can take that to the Talk page of the article. I haven't seen evidence of the existence of a reliable source being used as a "trump card" for inclusion on article talk pages.--greenrd (talk) 17:43, 8 February 2017 (UTC)
The problem with WP:V#Verifiability does not guarantee inclusion is that it is way down the page, with no hint of that in the nutshell or the lede. (Which I suspect is as far as many editors ever read.) And I have seen the position taken that verifiability is sufficient for inclusion, making it an uphill battle to keep out "interesting" but non-useful bits of clutter. Perhaps it should be mentioned at the top?
I don't see WP:TRIVIA as really helpful here, as it focuses on lists and sections of trivia, not the some what random inclusion of shiny bits. Likewise for WP:DUE (a.k.a. WP:WEIGHT): that is more about "balancing" different points of view. And perhaps I wasn't clear enough about my third class: I have in mind where someone wants to give some particular detail or sub-sub-topic more coverage (and thus more prominence) than can be given to similar details. If you think of an article's outline as a tree diagram, this would be about all of the leaf nodes having similar treatment. ~ J. Johnson (JJ) (talk) 23:31, 8 February 2017 (UTC)
I have seen the position taken that verifiability is sufficient for inclusion - me too, but the answer already lies in WP:ONUS and WP:INDISCRIMINATE or WP:NOTEVERYTHING, which clearly state that verifiability is not sufficient. BrightRoundCircle (talk) 01:22, 9 February 2017 (UTC)
I think that that argument is implicitly made very often. The common statements along the line of saying only "Undo removal of sourced material" implicitly make that argument. North8000 (talk) 13:54, 9 February 2017 (UTC)
That's true, and I've undid edits in those words exactly. As Greenrd said, verifiability is assumed to be good enough unless shown otherwise. Also, thanks for the comment on merely-verifiable listings. I was hoping the proposal is polished enough, but you pointed out a giant flaw that needs to be worked out. BrightRoundCircle (talk) 21:00, 9 February 2017 (UTC)
Thanks, and thanks for your efforts in these areas. North8000 (talk) 00:18, 10 February 2017 (UTC)

We have the essay Wikipedia:Coatrack articles, about articles that (for whatever reason) devote more space to the subject's background information than to the subject itself. Cambalachero (talk) 14:25, 9 February 2017 (UTC)

"Clearly state"? It may be clear enough for you, and clear enough for me, but is it obvious enough for a large fraction of editors? (See also: WP:It's obvious.) ~ J. Johnson (JJ) (talk) 23:59, 9 February 2017 (UTC)
I'm not saying all editors are familiar with every detail of every Wikipedia policy; I'm saying that in case of a content dispute as outlined by you above, the wording is clear. Would Wikipedia be a better encyclopedia if we put "Verifiability does not guarantee inclusion" at the top of WP:V? Maybe. I'll certainly support this change. BrightRoundCircle (talk) 01:31, 10 February 2017 (UTC)
"Guarantee" bothers me. Too much of the "formal assurance", and of an alternative in lieu of something else. I would prefer something more like "verifiability is required, but not in itself sufficient, for inclusion." ~ J. Johnson (JJ) (talk) 22:56, 10 February 2017 (UTC)
And there's my ole idea: "Verifiability is a requirement for inclusion, not a reason for inclusion". North8000 (talk) 23:49, 10 February 2017 (UTC)
I also like J Johnson's idea. North8000 (talk) 11:38, 11 February 2017 (UTC)
Whereas I rather like your wording. Perhaps we can find a good synthesis formulation? ~ J. Johnson (JJ) (talk) 21:31, 12 February 2017 (UTC)

I support J Johnson's wording as I think it gets the message across a bit clearer. I think we need it. Doug Weller talk 15:56, 12 February 2017 (UTC)

I'd support it, is this for the lead? Alanscottwalker (talk) 16:18, 12 February 2017 (UTC)

Yes. The underlying issue is not that the point is not addressed, but that it is (evidently) not prominent enough. If we come up with a good formulation I'll open an RfC for including it. ~ J. Johnson (JJ) (talk) 21:36, 12 February 2017 (UTC)
I support J. Johnsons and like mine even better. Even though both are just a re-statment of current policy and guidelines, I think that mine goes a step farther than theirs. "Not a guarantee" rules out only the most extreme case, a "gaurantee" It leaves it open to other current myths which might be that verifiability is a strong argument for inclusion. Mine rules out those myths as well. But I also suppport inclusion of J Johnson's version. 22:19, 12 February 2017 (UTC)North8000 (talk)
I'm partial to the wording in WP:NOTEVERYTHING and WP:INDISCRIMINATE: "Verifiable and sourced statements should be treated with appropriate weight"; "merely being true, or even verifiable, does not automatically make something suitable for inclusion." But if you don't actually change the wording and just move WP:ONUS higher up, directly below WP:PROVEIT in its own level-2 heading section, then the change is, in my eyes, not substantial enough to require an RfC, and a regular talk page consensus is enough. BrightRoundCircle (talk) 23:32, 13 February 2017 (UTC)
Another good idea that I support. North8000 (talk) 23:36, 13 February 2017 (UTC)
Same here. Staszek Lem (talk) 18:44, 14 February 2017 (UTC)

Proposal

  Done This and the additional proposals (below) have been implemented, as subsequently modified. ~ J. Johnson (JJ) (talk) 21:33, 28 February 2017 (UTC)

I think the WP:ONUS section (with the explanation and related links) is fine where it is, under "Other issues"; we just need to put in a high-level reminder. I propose that we add, following the third sentence of the lead ("Even if you're sure something is true, it must be verifiable before you can add it."), this text: "However, being verifiable is not sufficient for including something."

Do we need an RfC to do this, or is consensus here sufficient? — Preceding unsigned comment added by J. Johnson (talkcontribs) 08:05, February 15, 2017 (UTC)

  • Support. The change is minor enough not to require an RfC, and everybody in the discussion above agree that it wouldn't hurt for this policy to be more visible. Perhaps wikilink "not sufficient for inclusion" to WP:ONUS. BrightRoundCircle (talk) 09:06, 15 February 2017 (UTC)
Yes, a link would be good. As to an RfC: we're not changing policy, just clarifying it. ~ J. Johnson (JJ) (talk) 19:20, 15 February 2017 (UTC)
  • Comment Instead of "not" I'd rather have "not necessarily" in which case I'd agree.--Kmhkmh (talk) 14:02, 15 February 2017 (UTC)
  • Support, with the two amendment by BrightRoundCircle & Kmhkmh. Staszek Lem (talk) 18:32, 15 February 2017 (UTC)
  • Support as written. I don't see the need for confusing it by adding "not necessarily" Particularly since what we're really talking about whether or not verifiability is sufficient to force inclusion. North8000 (talk) 18:54, 15 February 2017 (UTC)

So we have an unresolved issue. One way of looking at this is: should material be allowed simply because someone said it, when there is no other reason whatsoever for including it? My idea is that the bar for including something is so low (given that it is verifiable) as it would not preclude anything worthwhile, while any material failing that – especially if there is any controversy about it – really ought not to be included. To allow "not necessarily" seems to leave open this door that verifiability alone is sufficient for inclusion. Alternately, I keep trying to think of examples of worthwhile material for which there is no reason for including, and keep hanging up on "worthwhile" being a reason for inclusion. Perhaps someone else can come up with some examples? ~ J. Johnson (JJ) (talk) 19:45, 15 February 2017 (UTC)

Keep in mind that this needs to govern all articles and we have areas or subjects when it is reasonable to assume that some information bit is generally of interest if it is verifiable even if secondary sources might not exist or fully cover it. For instance statistical or geographical data like the population of town, the altidude of a mountain, date of birth of a person, etc. . Another aspect that "worthwhile" for inclusion also depends on the total information being available on some (relevant) topic in the first place. For topic well covered with secondary sources, one usually would not include any material that is treated in those secondary sources and that's where your suggested formulation applies. But for (relevant) topics with little information in secondary, one might be more likely to include additional material that is verifiable "only" as long as it makes common sense as content.--Kmhkmh (talk) 20:10, 15 February 2017 (UTC)
I think that if you take the proposal literally,one will miss what this is really operatively about. It goes without saying that there needs to be some reason for inclusion beyond meeting wp:ver, but that per se is a non-issue. Nobody is knocking stuff out saying "no reason for inclusion given". What is an issue a gazillion times per day is someone essentially saying that meeting wp:VER is a strong-enough reason to let them force inclusion. IMO the graceful way to say otherwise is my original proposal..."verifiability is a requirement for inclusion, not a reason for inclusion." I support J. Johnsons proposal (without the "not necessarily") because it helps in that direction. Sincerely, North8000 (talk) 21:11, 15 February 2017 (UTC)
I am not adverse to replacing the proposed text with: "But while verifiability is a requirement for inclusion, it is not a reason for inclusion." But my concern there is that some editors would insist that they don't need any reason for inclusion.
Kmhkmh: isn't "generally of interest" a sufficient reason for inclusion? ~ J. Johnson (JJ) (talk) 23:21, 15 February 2017 (UTC)
Respectfully, I think that that misses a core point. The only time that this comes into play is on challenged material when the inserter wants to use meeting wp:ver as a force for inclusion. North8000 (talk) 12:49, 16 February 2017 (UTC)
Sort of, the problem here is that "generally of interest" in such cases is based on editorial judgement/discretion rather than an external criteria. Anyhow in doubt I can live your suggestion, northstar's suggestion, mine or the current version - they all work, just having a slightly different emphasis or visibility.--Kmhkmh (talk) 05:15, 16 February 2017 (UTC)
Ultimately everything comes down to an "editorial judgment", but having a few guidelines provide a basis for resolving clashing judgments. I note that the examples you provided earlier (populations of towns, altitude of mountains, birthdates) are all pretty standard descriptors, so even if someone else thought they are not "generally interesting" there is still a reason besides mere verifiability for including them. And if there are any problems we can always tweak the wording some more.
I was mulling over some possible refinements of the proposed text, then realized it doesn't have to be perfect, just good enough. And I think it is. ~ J. Johnson (JJ) (talk) 06:12, 16 February 2017 (UTC)
  • Oppose. The ONUS section is plenty clear. Adding an additional reminder would be superfluous. Someguy1221 (talk) 23:28, 15 February 2017 (UTC)
    It can hardly be a re-minder when a substantial fraction of editors apparently don't read far enough to get to ONUS in the first place. What is proposed is rather like a scope statement (common in some forms of writing) that the topic goes this far, but not that far. It is a sparse statement ("verifiability is not sufficient for inclusion") which is expanded and explained way down the page. Even if you don't feel it is necessary, it doesn't hurt anything to have it. ~ J. Johnson (JJ) (talk) 06:27, 16 February 2017 (UTC)
  • Support - a good lead highlights the important points that are discussed in more detail later in the policy. ONUS is an important point... and so should be highlighted in the lead. Blueboar (talk) 13:01, 16 February 2017 (UTC)
  • Support, generally along the same lines as wp:lead. Alanscottwalker (talk) 13:49, 16 February 2017 (UTC)
  • I'm not sure about this. Actually, I'm not sure about the value of ONUS, as written. I (as many of you know) tend to hang around the medical articles. We've got some fairly stringent rules about ideal sourcing, plus some editors who like to enforce them. One unintentional result is that some of the articles end up being – well, boring. Common cold gives an analysis of the current state of scientific evidence for a wide variety of treatments (including honey, Vitamin C, and Zinc lozenges, all of which it calls "Alternative medicine"), but it doesn't mention things like chicken soup or the famous old hot toddy. It doesn't even say that "boosting your immune system" is exactly the wrong approach (all the cold symptoms that people dislike are directly caused by your immune system, not by the viruses), so unless you happen to understand the one-paragraph ==Pathophysiology== section and make the connection yourself, then you'll think that Echinacea is a reasonable idea, even if the article hints that it might not work.
    I could blather on, but my basic reaction is this: Emphasizing this is going to be used to argue for the removal of anything that might be fun or interesting from articles, even if it's just a brief mention towards the end of an article, and I don't think that encouraging even more boringness is appropriate or desirable from the POV of readers. WhatamIdoing (talk) 05:53, 17 February 2017 (UTC)
I believe removal of "interesting" stuff – distinguished from mere trivia – is usually done on the basis of "unencyclopedic tone". And I would argue that articles can (and should) be written to be interesting ("tasty") without needing a bunch of pointless trivia. But the bottom line here is that material included for no reason other than "it's verifiable!" really has no reason for inclusion at all; such material should be removed. As for removing any material that is in any way relevant, or even just illustrative, I think you need not be concerned. As North8000 has observed: "The only time that this comes into play is on challenged material when the inserter wants to use meeting wp:ver as a force for inclusion." That an article is boring is beyond the scope of WP:V, and in any event the proper resolution for "boringness" is a proper rewrite, not adding sugar.
I think the problem with the WP:ONUS paragraph is that some editors slide over the first sentence, and then assert mere verifiabilty as meeting the onus for inclusion. I am contemplating a corrective for this. ~ J. Johnson (JJ) (talk) 22:45, 17 February 2017 (UTC)
WhatamIdoing: If you can't convince others that something is "interesting", perhaps it's not, and importantly verfiability can't save it. -- Alanscottwalker (talk) 11:01, 20 February 2017 (UTC)
No, it means that I'm dealing with WPMED folks, some of whom would rather have an article that contains nothing except an analysis of scientific evidence for and against current treatments. We just spent 7,000 words talking about whether a source that is actually relied upon by professionals around the world is "reliable", right? And several people left that discussion with the belief that the current regulatory status for billion-dollar project was unworthy of inclusion.
It's not a question of tone: Neither "Chicken soup is a popular home remedy for the common cold" nor "BigPharma began a Phase 2 clinical trial of Product X for Condition Y in 2016" have any tone problems. The problem is that providing accurate, basic information about experimental treatments and popular self-care methods is not what some editors think belongs in any encyclopedia – perhaps especially not in any encyclopedia that their patients (or any of their colleagues who don't yet believe that scientific evidence is the sum total of medicine's purpose) might read. WhatamIdoing (talk) 02:24, 21 February 2017 (UTC)
Okay, I think we can allow that there are problems in WPMED. But just how does the language here, either current or proposed, affect those problems? You are saying that the problem is that certain information "is not what some editors think belongs in any encyclopedia". That sounds more like an argument about the nature of the information, or perhaps the basis for incluson, neither of which is really a verifiability issue. When you, or any other editors, think something should be included, does that thinking have any kind of basis? I.e., a reason? (I was thinking of qualifying that with "other than WP:ILIKEIT", but heck, even that is a reason.)
Note that when I suggest that inclusion should be based on reasons other than WP:V, I am not suggesting we are limited to AMA approved "reasons". I am saying that if mere verifiability is the sole argument ("reason") for including something, then there really isn't any reason at all. And I think WP:V would be better off it wasn't used in inclusion battles. ~ J. Johnson (JJ) (talk) 00:49, 22 February 2017 (UTC)
WhatamIdoing: can you live with these proposals? ~ J. Johnson (JJ) (talk) 22:22, 24 February 2017 (UTC)
  • Support -- a step in the right direction. K.e.coffman (talk) 02:06, 21 February 2017 (UTC)

Additional proposals

I have boldly added a WP:VNOTSUFF shortcut to the "onus" section. In line with the forgoing disucssion I propose additional adjustments to the title and language of that section, as follows:

Proposal 2a: Change title: "Verifiability does not guarantee alone is not sufficient for inclusion
Proposal 2b: Change first sentence: "While information content must be verifiable in order to be included in an article, verifiability alone is not sufficient for inclusion, and does not guarantee that any content this does not mean that all verifiable information must be included in an article."

Comments? ~ J. Johnson (JJ) (talk) 23:32, 17 February 2017 (UTC)

Support. Like the previous proposal, a small step in the right direction. Wording that addresses the actual issue (meeting wp:ver is a requirement for inclusion, not a force for inclusion of contested material) would be even better. North8000 (talk) 11:52, 20 February 2017 (UTC)
Support "2". Easier to understand this way. Weak-oppose on adding another shortcut, "onus" shorter and more memorable than "v not suff". BrightRoundCircle (talk) 11:13, 23 February 2017 (UTC)
Support Same as before, a step in the right direction. As before, I still think "requirement, but not a reason" type wording would be more precise and useful.North8000 (talk) 12:08, 23 February 2017 (UTC)
I have been thinking of something more like: "While verifiability is required for including content, verifiability alone is not a sufficient reason for inclusion, and does not guarantee that any content must be included in an article." But having proposed a specific change, I don't like to be modifying it. Well, I don't believe anyone here is hung-up on the specifics, and I would just slap something in. Except that I would like to hear from WhatamIdoing that she can live with any of this. ~ J. Johnson (JJ) (talk) 20:30, 23 February 2017 (UTC)
I support that version as well. But I would like to do a logical dissection that will show why I think we've accidentally made this unnecessarily complicated, and to argue why my "Verifiability is a requirement for inclusion, not a reason for inclusion." is a better approach. First, let's note that I believe that this thread is to address a narrow but common situation. The situation always involves an at least minor conflict on whether or not to include the material. And then someone mis-uses wp:ver, in essence stating that meeting wp:ver is a significant (or even sufficient) reason for inclusion. Usually this is implicitly or explicitly saying that meeting wp:ver alone is a significant enough reason for inclusion to override some other reasons for removing it. (although they usually they don't go so far as claiming that it overrides other clear WP restrictions against inclusion). The description covering all of these situations is that the person is arguing that wp:ver is a reason for inclusion.
Whether or not some instance of meeting wp:ver is "sufficient" for inclusion (or similar wording in some of the proposals) is getting into a widely varying complex situation. To start with, an inclusion must comply with any clear WP restrictions on content. Beyond that, in some cases, like a one-editor obscure article where nobody is contesting the inclusion, nothing else is needed to include it. At the other end of the spectrum, where it is in the middle of a hot dispute, a requirement for inclusion might end up being obtaining a consensus in an RFC. So if you start trying to talk about "sufficiency" for inclusion, you are wading into very broad complex area that is not the purview of wp:ver or any other single policy. So, in my view, what comes out of this thread should simply address the mis-use of wp:ver, not wade into that huge complex area of sufficiency for inclusion. IMO the way to say this is "Meeting wp:ver is a requirement for inclusion, not a reason for inclusion". BTW, this is a summary of what is and isn't in current policies....it handles a widespread myth/mis-use, but does not represent any new policy or change any policy. North8000 (talk) 21:17, 26 February 2017 (UTC)
Okay. How about if my proposed addition is replaced with: "However, while verifiability is required for including something, it is not a reason for inclusion."
For my additional proposals we can replace "not sufficient for" with "not a reason for". (Though in the text I'm still inclined towards "not a sufficient reason for".) I think most everyone here is generally agreeable, and not going to sweat the details. If you and I agree on the details I'll make those changes. ~ J. Johnson (JJ) (talk) 08:31, 27 February 2017 (UTC)
Support!!!!!. The exclamation marks are for the last proposal exactly as written. And again,it does not represent any new policy or change any policy. North8000 (talk) 12:47, 27 February 2017 (UTC)

Okay, I've done did it: the changes are in, more or less as discussed; let's see how they fit. We can still discuss them if anyone has an concerns or questions, or just a comment. But hopefully we have made one tiny lurch forward. ~ J. Johnson (JJ) (talk) 21:18, 27 February 2017 (UTC)

CoolNorth8000 (talk) 22:12, 27 February 2017 (UTC)
Let's see how it goes. If it doesn't work out, then we can try something else. (Thanks for the ping above, too.) WhatamIdoing (talk) 05:23, 2 March 2017 (UTC)

Contestation of consensus

[This section arises from S Marshall's contestation of the consensus reached above, and discussed at length below at #Back-and-forth about putting "sufficient" into the lede. ~ J. Johnson (JJ) (talk) 20:34, 28 March 2017 (UTC) ]

@S Marshall: There is consensus. Just because you missed the discussion doesn't mean the consensus is invalid. The "stable version" wording was made without discussion, and this consensus holds until a new consensus is reached. There is no policy that favors a "stable version". BrightRoundCircle (talk) 00:36, 23 March 2017 (UTC)

  • Well, as you don't have many edits to this talk page, I'm bound by WP:BITE, and I'm having trouble coming up with a diplomatic response. Could someone else please address BrightRoundCircle's, err, alternative facts here?—S Marshall T/C 01:28, 23 March 2017 (UTC)
I opened a section on this specifically below. Having worked with you in the past, I know that you are one of the most brilliant policy people in Wikipedia. If you look at the various logical "layers" involved (some of which I just detailed below), IMHO possibly you might think differently ? Sincerely, North8000 (talk) 01:39, 23 March 2017 (UTC)
@S Marshall: Kindly point out which policy says "stable version" overrides consensus and leave your high-and-mighty attitude elsewhere. "I missed that discussion" is also not a reason to undo consensus. You want changes, create a new consensus. BrightRoundCircle (talk) 10:38, 23 March 2017 (UTC)
  • Look, I've followed this talk page carefully for the last ten years and accumulated more than a thousand edits to it over that time. I inadvertently missed one discussion between what was, by the standards of substantial changes to WP:V, a minuscule number of editors a few weeks ago that led to some very problematic changes to the policy, and you basically just told me I was screwed. It wasn't a RFC, it wasn't closed by an independent editor, it didn't stay open for the requisite 30 days, so it's not an unchallengeable consensus that I have to accept. I'm entitled to revert it. The stable version of a policy is used until there's consensus otherwise: that's the editing policy. AND, your statement that "the stable version was made without discussion" is just ignorant and ridiculous. You didn't take part in the discussion, but literally every single word in this policy has been subject to close scrutiny from the community. You need to back off and allow the stable wording to be restored.—S Marshall T/C 13:18, 23 March 2017 (UTC)
I'm entitled to revert it - that's the point, you're not. If you missed a discussion, you're not entitled to revert its consensus. Note that WP:EP says "present consensus", not "stable version". The present consensus was reached in a discussion. There is no policy for reverting to a "stable version", only WP:CONSENSUS. You want a new consensus, create it. literally every single word in this policy has been subject to close scrutiny from the community - no. Much of Wikipedia policy has been written without discussion. The version of WP:ONUS you are attempting to restore has been discussed without an RfC or closure by an independent editor. It does not supersede the new consensus; the new consensus supersedes the old consensus according to Wikipedia policy. Don't like it? Create a newer consensus. Create an RfC. Have an independent admin close it. But don't revert because of "stable version" and "I missed that discussion" and don't tell me you're "entitled". BrightRoundCircle (talk) 16:55, 23 March 2017 (UTC)
The paragraph was introduced by @Blueboar: in this edit after the discussion which is now in the archives here. I'm going to tell you one last time: there's nothing that's still in this policy that hasn't been thoroughly discussed and closely scrutinised by the community.—S Marshall T/C 19:10, 23 March 2017 (UTC)
All your other assertions were wrong so I might as well show you this one's wrong too. But to recap:
  • You're not "entitled" to revert a consensus simply because you were absent from the discussion.
  • There is no policy for "stable version"; the only relevant policy is consensus.
  • "It wasn't a RFC, it wasn't closed by an independent editor" - neither was the previous consensus nor the "stable version". Not all discussion has to be an RFC and closed by an independent editor after 30 days; the previous consensus wasn't, that doesn't invalidate it. The previous discussion is more or less on the same scale of participation as the current discussion. If you want a newer consensus you are free to create one.
Now back to your last point: there's nothing that's still in this policy that hasn't been thoroughly discussed - but there is. For example this very large change, most of which persists through this day ("anyone can create a..."). You can verify in all the talk pages up to and around the date of the change that there was no discussion about this very large change. This example of consensus-by-editing with no discussion is not uncommon in the early stages of all Wikipedia policies, where someone made a good, thorough, well-worded contribution and nobody objected. While consensus-by-editing is great, it doesn't mean it was "thoroughly discussed". So I'd say you're wrong on all counts. Rudeness is not a substitute for having valid points. BrightRoundCircle (talk) 09:59, 24 March 2017 (UTC)

Correct: rudeness is not a substitute for having valid points. I couldn't agree more and I hope you'll desist.

You've made a recent, substantial edit to a core policy, without an RfC. I'm entitled to revert it if I dispute it. You've reverted it back; so we discuss. And everything in the policy has been thoroughly discussed and closely scrutinised by the community. I did not say this happened at the time the change was made.—S Marshall T/C 09:57, 28 March 2017 (UTC)

@S Marshall:: The rudeness here is predominantly, perhaps even entirely, on your account. E.g.: your insistence that you are entitled to revert, your adamancy in rejecting any contrary view, your impugning BRC's comments as "alternative facts", your telling others to back off, your imputation of rudeness, and your general non-cooperative attitude that amounts to incivility. This is becoming a behavioral problem that transcends the policy or process issues. ~ J. Johnson (JJ) (talk) 20:58, 28 March 2017 (UTC)
P.S.: I concur with the comments of North800 and BrightRoundCircle. ~ J. Johnson (JJ) (talk) 21:00, 28 March 2017 (UTC)
If you think I'm misbehaving, you're welcome to file for dispute resolution or on AN/I. Be my guest.—S Marshall T/C 21:30, 28 March 2017 (UTC)
Your behavior aside, your edit summaries and rationale (missed the discussion / stable version / no RfC) do not override consensus. None of this is "alternative facts" as you rudely suggest and your entire discussion method is to lob insults and avoid the core issue - which is, your revert reasons do not override consensus. BrightRoundCircle (talk) 13:19, 29 March 2017 (UTC)

Self-published or questionable sources as sources on themselves

I added an "and" to the list of criteria for self-published sources on themselves to clarify that this list is conjunctive, not disjunctive. I was immediately reverted on the grounds that "(1) it is disjunctive, (2) human language is not a programming language, lists do not need to specify and/or." First, the list is pretty clearly conjunctive: it would make no sense to argue you could use a self-published source if there was "no reasonable doubt as to its authenticity" but it failed the other criteria. Second, I'm not a programmer, but it seems silly to argue that adding an "and" for clarity is worthy of reversion simply because it arguably isn't a mandatory inclusion. Dyrnych (talk) 18:46, 2 April 2017 (UTC)

Okay. Bright☀ 19:03, 2 April 2017 (UTC)

Proposal regarding use of self-published social media as sole source of factual content at BLP and other articles

I invite @Blueboar: @S Marshall: @North8000: and any other interested editors return to the section "Please clarify—authoritative responses only (admins and experienced editors)", regarding the appropriate use of Tweets, Facebook, and other self-published sources, as sources to satisfy WP:VERIFY, and review that material. Then, summarise final opinions here. Include any thoughts on the interpretation of WP:ABOUTSELF as a possible intended statement of exception that allows editors to use self-published expressions about themselves as sources at WP (i.e., to self-report birth dates, places of residences, details about events in which they are involved, etc.). Cover, please, if it makes any difference to your conclusions, if the accounts that are the sources of the self-published communications are "verified accounts," or not. Please cite specific policy and guideline statements, as much as possible, to establish this or a different proposal position. Thank you. Le Prof Leprof 7272 (talk) 20:52, 23 March 2017 (UTC)

Proposal. That we agree, that both with respect to the spirit of the policies and guidelines, and with regard to specific statements, we are to exclude all self-published social media as sole source, regardless of social media account type, for BLP and other biographical articles, as well as articles about events, places, etc. (essentially all articles), except as a supplemental source placed in addition to a third-party (non-self-published) source, that unequivocally establishes the fact in question. Le Prof Leprof 7272 (talk) 20:52, 23 March 2017 (UTC)
The policy says Self-published and questionable sources may be used as sources of information about themselves and lists five caveats. While I think the first and last caveats need elaboration, I don't see how the policy excludes self-published social media that fits within the details and caveats. --Ronz (talk) 21:25, 23 March 2017 (UTC)
What precisely does "about themselves" mean in the policy? That orgs can make statements about themselves on FB, e.g. regarding an upcoming event? That a person can tweet their BD and we use it as source? One interpretation is that we can use an SM source to say that the source made the statement—for instance, that this alone is allowable, "Cristina Vee has tweeted that her birthday is... [cite tweet]" (versus "Cristina Vee's birthday is... [cite tweet]"). And what elaborations do you propose for the first and last caveats?
Extended content

Self-published and questionable sources may be used as sources of information about themselves, usually in articles about themselves or their activities, without the self-published source requirement that they be published experts in the field, so long as: 1. the material is neither unduly self-serving nor an exceptional claim; 2. it does not involve claims about third parties; 3. it does not involve claims about events not directly related to the source; 4. there is no reasonable doubt as to its authenticity; 5. the article is not based primarily on such sources. This policy also applies to material published by the subject on social networking websites such as Twitter, Tumblr, Reddit, and Facebook.

Note that cases of event participants tweeting about addition of others to an event appears to fall within the current Caveat 2 (collapsed above). And that the long history of published cases of obfuscation and dissembling about DOBs seems to fall under Caveat 1. Finally, one change that might be needed to C5, is to add "section"—that a sole source for a section or article... (As stated above, despite the firm Roberts Rules-type motion, I less have a horse in the race, than I wish to have clarity to allow me to engage others.) Le Prof Leprof 7272 (talk) 21:33, 23 March 2017 (UTC)
Why would the author's social media have an extraordinary requirement of requiring the third-party source? It should be treated as with any other primary source whether it be a video statement, a published written statement, or an autobiography. AngusWOOF (barksniff) 00:39, 24 March 2017 (UTC)
Because it is self-report and a primary source? And in some cases, has a pervasive chance of COI? In many cases the author might be the best source, but it doesn't even amount to journalism. A journalist saying Joe Blow was partying a certain night is a rather different matter than hearing Joe Blow saying that. I'd say that self-report always requires extra care in vetting. ~ J. Johnson (JJ) (talk) 23:19, 28 March 2017 (UTC)
I'm not saying it shouldn't have care in vetting, but it still should have equal vetting as the published written statement, interview statement, or autobiography.AngusWOOF (barksniff) 19:36, 2 April 2017 (UTC)
I don't agree you can merely equate those things. There is a difference in review before publication. Alanscottwalker (talk) 19:54, 2 April 2017 (UTC)
How is there a difference in review? So the publisher's editor corrects the statements for spelling/grammar? They don't fact check what the person writes about themself, and the newspaper/magazine usually has an implied disclaimer that it's the person's writing and not their own editorship? AngusWOOF (barksniff) 20:20, 2 April 2017 (UTC)
How is there a difference in review? First because of the mode of production, longer formal written statements are longer formal written statements. Second, when there is a publisher, they regularly do fact check, virtually always clear legal, as well as shape prose. Alanscottwalker (talk) 20:25, 2 April 2017 (UTC)
If that's the case, would published statements about birth dates and residences, if put in writing by autobiographies or press releases, be considered a better source than the tweets and website posts? AngusWOOF (barksniff) 20:52, 2 April 2017 (UTC)

Getting that prescriptive to me sounds like wp:creep. A better change (or tweak) of policy would be to say that if someone expresses a true concern about the veracity of the statement which cites it (and thus also the strength of the sourcing) that stronger sourcing should be provided. Maybe this little area could be a start for that much-needed change. North8000 (talk) 11:23, 24 March 2017 (UTC)

I concur that the proposed language is overly prescriptive.—S Marshall T/C 14:09, 24 March 2017 (UTC)
"About themselves"? The tweet is sufficient to back up the assertion that "@Some_twit said X",[ref tweet] not sufficient to assert "@Some_twit is the current reincarnation of @Prior_twit".[ref tweet] (my emphasis for discussion only). Otherwise we'll have an endless issue of aging actors and models youthifying themselves, self-appointed living "saints", etc. The SPS (however published) can only be considered to support assertions that the source says what it says. It has no value in assessment of its own accuracy and should never be used to put something in the voice of the encyclopedia rather than the SPS voice. LeadSongDog come howl! 20:10, 24 March 2017 (UTC)
LeadSongDog, I would contest a birth year that's only supported by a tweet as "unduly self-serving", because yes, actors often withhold their age even on their websites, but the case here is more about information such as month and day, which most actors do not mess around with, or hometown and current residence. If the actor's social media tends to be full of fictional statements, half-truths though or "crying wolf" statements, or seems to be a confusing persona gimmick as with Andy Kaufman / Latka Gravas, or Rodney Dangerfield the entire channel could be contested, but it should be considered on a case-by-case basis. AngusWOOF (barksniff) 16:32, 28 March 2017 (UTC)
Perhaps I should have said that the wording "About themselves" is confusing. The referent "themselves" is the SPS itself, not the author-publisher of that SPS. An SPS, whether a tweet or a patent, is useful only to verify its own wording, not the ideas asserted in those words. LeadSongDog come howl! 18:16, 28 March 2017 (UTC)
  • Sounds like a terrible source but just out of curiosity, how do you cite a tweet? Alanscottwalker (talk) 20:27, 25 March 2017 (UTC)
Alanscottwalker, check out {{cite tweet}} AngusWOOF (barksniff) 16:32, 28 March 2017 (UTC)

`

Back-and-forth about putting "sufficient" into the lede

In the back and forth in essence on putting "sufficient" in I would like to recap a few of the reasons for keeping it out:

  • First, WP:ver in essence imposes a requirement for inclusion, nothing more. Nowhere does it say anything about meeting wp:ver being a full or partial reason for inclusion. The lede of the policy should be a summary of the policy. "Not sufficient" clearly implies that, per wp:ver, meeting wp:ver is, to a partial extent, a "reason" for inclusion. So, including "not sufficient" in the lead is in conflict with the body of the policy.
  • Second, as detailed at length in the discussion on this, a common problem (rooted in a false false urban legend in Wikipedia) is mis-use or mis-reading of this policy to claim that meeting wp:ver is a force or reason for inclusion. Nobody makes the claim that it is an absolute mandate for inclusion, but it is quite common to make the false claim that it is a strong force or reason for inclusion. For example, every time an editor says the ubiquitous "undo removal of sourced material" in an edit summary they are in essence making this claim. Flatly saying that meeting wp:ver is not a reason for inclusion in the lede is (unlike "not sufficient") an accurate summary of of the body of the policy, and directly addresses the false urban legend.
  • Third, IMHO many folks (including, going by memory, one or more current or ex Arbcom members) have said that Wikipedia has an overly byzantine set of confusing and overlapping / conflicting policies. This is a perfect example of sitting on the edge of that precipice, and avoiding the abyss. Regarding content, WP:VER sets a requirement for inclusion, nothing more unless we let it be a mess. It should not try to wade into the immensely complicated question of "sufficiency" for inclusion. To start with "sufficient" requires not being in violation of dozens of policies and core guidelines. If those are complied with, there is an immensely varying additional bar for inclusion. At the low end might be an uncontested inclusion at an obscure article where the additional bar is essentially zero. At the other end of a spectrum, the additional bar for inclusion (on a contested major item) might be getting a strong consensus in an RFC. WP:VER should stick to imposing the verifiability requirement, not start reflecting on or overlapping with the hugely complex area of what constitutes sufficiency for inclusion.
  • Fourth, for new timid editors, this falsely implies that there is some unknown-to-them wiki-rule about needing a compelling reason to include something. This imposes a false hurdle, conflicts with new editor objectives (and wp:bold)

Sincerely, North8000 (talk) 01:16, 23 March 2017 (UTC)

  • If a thing is verifiable (by way of an inline citation to a reliable, published source) then it probably belongs somewhere in the encyclopaedia. Wouldn't you agree?
    I feel that yes, verifiability should be a virtual guarantee of inclusion, provided the source is properly reliable. My preferred formulation for this part would contain the following sequence of ideas:-
    (1) Everything in our encyclopaedia must be verifiable. (2) Everything verifiable should be somewhere in the encyclopaedia; provided that (a) its placement is appropriate, (b) the prominence it's given is proportional to its importance, and (c) only the mainstream academic view is given in Wikipedia's voice -- non-mainstream positions should have in-text attribution (not "the moon is made of green cheese" but "according to XYZ, the moon is made of green cheese"). (3) In case of dispute about whether the content is appropriate in any given article, then (a) disputed text should be omitted from the article unless there is a consensus to include it, (b) if there is consensus to include it then it should be given low prominence unless there is a consensus to give it high prominence, and (c) if there is consensus to give it high prominence then it should be given in-text attribution unless there is a consensus not to.—S Marshall T/C 01:47, 23 March 2017 (UTC)
I absolutely do NOT agree that everything verifiable "probably belongs somewhere in the encyclopaedia". Your position that "verifiability should be a virtual guarantee of inclusion" is utterly ABSURD, and shows a poor grasp of the implications. Even if EVERYTHING (do you fully grasp the TOTALITY of that concept?) is limited to reliable sources (with the usual rather low bar for "reliable") there are (just in the sciences) thousands of journals with tens to hundreds of issues, and my mind balks at how much citeable content could be mined from any given issue. And then there are newspapers and magazines galore, and I don't know how many books – millions? – just chock full of quotable content that, under your concept of the matter, are guaranteed inclusion in Wikipedia? Encyclopedias, including Wikipedia, are supposed to be distillations of human knowledge, not an echo of everything that has been said on every possible topic. ~ J. Johnson (JJ) (talk) 08:08, 23 March 2017 (UTC)
J. Johnson is correct... a fact may be verifiable, and yet too trivial to be worth mentioning in any article in Wikipedia... an opinion may be verifiable and yet too fringe to be worth mentioning in any article in Wikipedia. Verifiability does not guarantee inclusion. Blueboar (talk) 11:12, 23 March 2017 (UTC)
An example would help here. Please show me an example of an edit that introduces information that's verifiable by means of an inline citation to a proper WP:RS, when that information doesn't belong anywhere on the encyclopaedia.—S Marshall T/C 13:21, 23 March 2017 (UTC)
Is that question helpful. No one is editing the entire encyclopedia, when these issues come up. They are editing a particular article. They may suggest another article (if they know of one, but what they do know is, 'it does not belong, here') but then the discussion moves to that other article (where the editors of that article decide), and so on. Alanscottwalker (talk) 14:10, 23 March 2017 (UTC)
The question may well not be helpful to you. An answer would certainly help me.—S Marshall T/C 14:26, 23 March 2017 (UTC)
But, how do you expect anyone to know, when 1) it is context specific, and 2) there are so many unknowns (as in articles) for each individual editor. But I will try to at least meet you part way: "The cat got out of the tree" (undoubtedly there are RS for that), does it belong anywhere. -- Alanscottwalker (talk) 14:51, 23 March 2017 (UTC)
If these edits adding verifiable, reliably sourced content that doesn't belong anywhere in the encyclopaedia really do happen, then someone should be able to produce a diff. Wouldn't you say?—S Marshall T/C 18:45, 23 March 2017 (UTC)
No. I would not. The question makes no sense to me. I don't expect editors to know all the articles, (let alone its paragraphs, clauses. content, and subjects) on Wikipedia, but I do expect them to actually edit (which regularly means cutting what is actually before them to edit). Alanscottwalker (talk) 19:13, 23 March 2017 (UTC)
Sure. And what experience of actually editing tells me, is that it's incredibly rare (if it happens at all) for someone to add verifiable content that doesn't belong anywhere on the encyclopaedia. I'm much more concerned about the other side of the coin, actually -- editors who want to remove verifiable content. In particular, I'm worried about editors who're advancing the best interests of their nationality, religion, preferred conspiracy theory or favourite brand of snake oil (delete as appropriate) and would like to remove information that disfavours it. The current wording is a bit of a gift to them. But before we get to that, let's establish that there really are edits that add verifiable information that doesn't belong anywhere on the encyclopaedia. At the moment I think that objection's hypothetical and rather forced.—S Marshall T/C 20:24, 23 March 2017 (UTC)
Your criterion of "information [that] doesn't belong anywhere on the encyclopaedia" is more absurdity. You have not specifed the nature or basis of "belong", and having it contingent on a universal disproof makes it effectively undecidable.
It seems that what really worries you is editors removing information that disfavours certain topics. If the "current wording is a bit of a gift to them" it is undoubtedly because there is no other reason for including that information. Or possibly you are not sufficiently skillful at adducing such reasons, but either way your argument begins to sound like special pleading. ~ J. Johnson (JJ) (talk) 04:33, 24 March 2017 (UTC)

No, look, it's not absurd. It's very simple and straightforward. I've just gone through Special:Recent changes trying to find an example edit, and it was impossible for me, because in the samples I checked, every piece of information that anyone added which is backed up with an inline citation to a genuinely reliable source was appropriate. So I resorted to an example from a controversial article that I actually edit: Electronic cigarette. In my opinion, that article is full of content which is reliably-sourced, but doesn't belong there. A lot of it is simply at an inappropriate level of detail, belonging in Regulation of electronic cigarettes, Safety of electronic cigarettes or Electronic cigarette aerosol and e-liquid. To that article, I've never seen someone add anything that doesn't belong anywhere.

You're right to say that removing information that disfavours certain topics is a big concern of mine. I come across this a lot when I close RfCs, where I've found for example people wanting to remove the term "pseudoscientific" when associated with intelligent design (diff); people wanting to remove the term "pseudomedicine" when associated with homeopathy and other kinds of snake oil (diff), or people wanting to remove the words "Pakistani victory" from Battle of Chawinda (diff). These are not constructive edits, and I'm happy to say that when WP:V says Verifiability is not a reason for inclusion, we're inappropriately handing free ammunition to editors who want to advance a promotional, nationalistic or pseudoscientific agenda in this way.

Still looking for diffs that support your position.—S Marshall T/C 11:24, 24 March 2017 (UTC).

Your diffs are not examples of contested edits we can scrutinize, but your own 600+ word RfC closing statements. How about a link to (say) where someone removed "pseudoscientific" (or anything) from Intelligent design on the basis is was not verifiable? As to edits that support "my position": it appears you have completely missed what my position is. In effect, you want a diff of an edit that, incidentally!, demonstrates that some thing does not anywhere in the entire encyclopedia. As I said before: absurd. ~ J. Johnson (JJ) (talk) 00:16, 25 March 2017 (UTC)
There must be some other reason for including those terms without merely relying on saying that they meet the verifiability requirement. North8000 (talk) 11:41, 24 March 2017 (UTC)
Right, and that's the key point. As you correctly say, what this ought to lead to is a discussion of the other reasons for including this information. If the policy went back to saying "Verifiability is not a guarantee of inclusion", which is what the policy rightly ought to say and has said for many years, then it would naturally lead to a discussion of other reasons for including the information. But at the moment it says "Verifiability is not a reason for inclusion", and that is a clear, unqualified licence to remove verifiable information from the encyclopaedia. Editors who want to advance an agenda will find this wording very convenient.—S Marshall T/C 14:07, 24 March 2017 (UTC)
That is part of what I do not get about SMarshall's argument: something has a source, yes, the next basic question is, so? We are aiming to present a coherent tertiary article -- no one can possibly do that without cutting from the blobulous mass of 'things people say.' Alanscottwalker (talk) 12:49, 24 March 2017 (UTC)
  • Let me try saying it differently. For example, Editor A inserts a, er, blobulous piece of information into an article and Editor B wants to remove it. What should Editor B say? Should he say, "That information doesn't belong there" and remove it? Or should he say, "Thanks for adding verifiable information to our encyclopaedia. I think it would fit better in another place such as X, Y or Z, though, so I've removed it from here." Basically where I'm coming from is, I'm trying to find words that promote a more helpful and supportive approach to editors who add verifiable content.—S Marshall T/C 14:07, 24 March 2017 (UTC)
I think that I understand S Marshall's point, although I don't agree that it should determine the result of this particular sentence. Ironically and unfortunately, we are both rightly concerned about how a policy might get mis-used. I'm more concerned (and feel that it is much more frequent) about the policy being used as a force for inclusion. Most commonly as a way to POV an article. S Marshall is concerned that the current wording might be mis-used as a way to remove material that belongs in the article. IMHO the current wording is logically neutral on the topic, and a logically accurate summary of the policy, whereas IMHO adding "sufficient" implies things which are not logically neutral and which aren't in the policy. Perhaps we could clarify so that it can't get mis-used in either direction. North8000 (talk) 15:43, 24 March 2017 (UTC)
Well, let's agree both ways. Firstly, that this is a Wikipedia policy and not Semantics 101, so we don't need to get into "necessary and sufficient"; and secondly, that "Verifiability is not a reason for inclusion" should be changed back to "Verifiability is not a guarantee of inclusion" (or, preferably because more succinct, to "Verifiability does not guarantee inclusion"). Once we've got that sorted, we can get on with discussing the difference between "information" and "content" and why this policy is about the former rather than the latter.—S Marshall T/C 17:00, 24 March 2017 (UTC)
No. "First do what I say then we'll discuss it" is not going to work, particularly because everybody else in this discussion does not agree with that wording. BrightRoundCircle (talk) 18:53, 24 March 2017 (UTC)
If that's the case then we'll go to full RfC, but you do need to let them speak for themselves about whether they agree with it.—S Marshall T/C 19:00, 24 March 2017 (UTC)
Could you clarify whether you are advocating the "which may be influenced by" end of the sentence which was in your edit? IMHO that was particularly problematic with respect to 3 of the 4 reasons that I described above. Sincerely, North8000 (talk) 21:16, 24 March 2017 (UTC)
I think "which may be influenced by" is needless.—S Marshall T/C 21:35, 24 March 2017 (UTC)
Stuart, has it occurred to you that anyone wanting "to advance a promotional, nationalistic or pseudoscientific agenda" might also add content that (like the content you would retain) has no basis for inclusion other than it is verifiable? And that, under your conception that mere verifiability guarantees inclusion, neither you nor anyone else would could remove it? What would you do then?
WP:V sets a minimum requirement for all content, regardless of whether it favors or disfavors any particular point of view. It does not take sides. ~ J. Johnson (JJ) (talk) 00:21, 25 March 2017 (UTC)
  • This statement that mere verifiability guarantees inclusion does not reflect what I said. It is a straw man. I did say that verifiability should be a virtual guarantee of inclusion somewhere in the encyclopaedia, iff the source is genuinely reliable; the qualifiers are important. I then presented a general challenge: can anyone produce a diff of someone adding verifiable information that doesn't belong anywhere in the encyclopaedia?
    The edit that I want to make to the policy is to restore the stable, longstanding, correct phrasing, or a slight variant ("Verifiability does not guarantee inclusion"); and I did produce several diffs of actual things that really happen on Wikipedia to explain why.—S Marshall T/C 00:49, 25 March 2017 (UTC)
Not a strawman argument at all. A virtual guarantee seems plainly and clearly a form of guarantee, and if you have some finer distinction here that is crucial to understanding your position it is well past time that you explained it. Qualify it as you will, unless you actually negate "guarantee" what we have is a "formal promise or assurance" that something will be included. And no matter you qualify it, you still have not addressed your absurd criterion of something "belonging" (however you define that) "[somewhere/anywhere] in the encyclopaedia".
The diffs you produced above are NOT "diffs of actual things that really happen on Wikipedia", they are your long-winded statements not worth the time to read them. If you have diffs of actual edits to articles where some material was removed for the sole reason of being unverifiable, please provide them in this discussion. ~ J. Johnson (JJ) (talk) 23:33, 25 March 2017 (UTC)

Force line break

Might as well mention that the wording was discussed and consensus has been reached, so until a new consensus is reached this version should remain. BrightRoundCircle (talk) 10:46, 23 March 2017 (UTC)

Agree. North8000 (talk) 12:28, 23 March 2017 (UTC)

@S Marshall: Addressing the elements in your 01:47 23 March post. On your first question, I would agree that the spirit of what you said is often true, but not always true. Beyond that I'd probably say pretty much the same things that Blueboar said regarding that question. Plus, even if it were categorically true, there is no practical usable policy statement that could be made regarding "belongs somewhere in Wikipedia". Policies inherently relate to particular instances of presence in a particular article. Regarding the last 80% of your post, I agree with all of it but IMHO it does not relate to adding "sufficient" to the WP:VER policy sentence in question. Respectfully North8000 (talk) 12:40, 23 March 2017 (UTC)

@S Marshall: I disagree that every piece of verifiable information belongs somewhere in Wikipedia. As an example, states have websites which are reliable, and which list every conceivable government official, including select board members, justices of the peace, notaries public, and animal control officers. The state websites are even secondary, in the sense that they are compiled from municipalities who report their elections and appointments to the state. Frequently the home addresses and phone numbers of these officers are listed. So you think all this belongs in Wikipedia? Jc3s5h (talk) 21:47, 24 March 2017 (UTC)

  • What I said is that I've never seen an edit that adds verifiable information that doesn't belong anywhere in the encyclopaedia. I'd welcome counterexamples with diffs.—S Marshall T/C 21:51, 24 March 2017 (UTC)
Apparently he does think so.
Stuart, your criterion of "verifiable information that doesn't belong anywhere in the encyclopaedia is (as I keep saying) ABSURD, because it demands a universal disproof. It is effectively impossible to prove that a given factoid does not belong at each and every point in our five million+plus articles.
We don't add material because it "belongs" (how?) somewhere in the encyclopedia, but because it supports something at a particular point in a particular article. Whether certain information might "belong" at some point has no bearing on whether it belongs at any other point. ~ J. Johnson (JJ) (talk) 00:42, 25 March 2017 (UTC)
I'm not going to give diffs, but there is an example everyone has seen of verifiable information that does not belong in Wikipedia. Namely, the silly items of trivia that people like to add to articles on actors and other famous people. They make no sense at all in any article except the one about that person, but experienced editors usually take them out of that article too. So the community standard is that they don't belong anywhere. Zerotalk 02:25, 25 March 2017 (UTC)
In fact there is an entire policy about verifiable information that doesn't belong on Wikipedia. It's called WP:NOT. BrightRoundCircle (talk) 13:29, 25 March 2017 (UTC)

I don't really think either wording is ideal. But the present version is actually wrong as stated. As S Marshall points out, of course verifiability is a reason for including something. Generally speaking, especially in areas where other policies come into play, it is certainly not a sufficient condition (that is, it's not guaranteed). But for the average boring Wikipedia article, about some former olympic champion for example, of course the verifiability reason is probably also sufficient for things like listing records and medals and so forth. Most of the content on Wikipedia is actually of this kind (perhaps thankfully), in spite of our collectively warped noticeboard-informed views. Sławomir Biały (talk) 19:29, 25 March 2017 (UTC)

So when we go to RfC

I suggest that the question is:-

Please support position #1 or position #2, or add your own alternative below.

  • Position 1: Verifiability is not a reason for inclusion (status quo based on recent talk page discussion);
  • Position 2: Verifiability is not a guarantee of inclusion (longstanding wording); policy would read like this.
  • Alternatives/threaded discussion.

Any objections?—S Marshall T/C 16:19, 25 March 2017 (UTC)

I object, if this is just a free-floating philosophy RfC. There are two places in the policy at present this is discussed, in the lead and in the ONUS section, so if we are to have an RfC it should be a proposal to those specific sentences, in context. Alanscottwalker (talk) 16:41, 25 March 2017 (UTC)
Oh, yes, it's meant to be specifically about wording. Question amended.—S Marshall T/C 17:31, 25 March 2017 (UTC)

Of course I'd rather not have an RFC and keep it how it currently is. But one suggestion....your proposal would have fewer problems if you left off the "which may be influenced by other policies and guidelines" North8000 (talk) 18:43, 25 March 2017 (UTC)

Not clear. Are you suggesting altering just the header or the entire paragraph? If the RfC is about changing the entire paragraph, you should include it in its entirety:


Verifiability is not a reason for inclusion

While content must be verifiable in order to be included in an article, verifiability alone is not a reason for inclusion, and does not guarantee that any content must be included in an article. Consensus may determine that certain information does not improve an article, and that it should be omitted or presented instead in a different article. The onus to achieve consensus for inclusion is on those seeking to include disputed content.


This makes it perfectly clear what the RfC is about. If you want to change just the header you should make that clear, too. BrightRoundCircle (talk) 19:49, 25 March 2017 (UTC)

  • I'm talking about either maintaining the status quo or restoring the version in the diff, and frankly, that's not hard to see from what I wrote.—S Marshall T/C 20:11, 25 March 2017 (UTC)
So, are you only talking about the lead? Alanscottwalker (talk) 20:15, 25 March 2017 (UTC)
This would be a lot easier if he'd just include the wording he desires in the RfC, instead of just one sentence. BrightRoundCircle (talk) 20:26, 25 March 2017 (UTC)
The choice is between the status quo (i.e. how the policy looks now) based on a recent, poorly-attended discussion on the talk page, and the long-term stable version to which I'd prefer to return, which looks like (this). Both options are visible in the diff that's in the question. Participants will need to click the diff, read, comprehend, and then respond.—S Marshall T/C 22:02, 25 March 2017 (UTC)

S Marshall, that sort of outlines it clearly, although it should be turned into neutral specific words. It's worth spending time to get an RFC right; so many doom themselves by wording / structural problems. But the previous version had so many issues that you might be shooting yourself in the foot. Perhaps we could work out something? (one version, or else two to decide between) North8000 (talk) 22:18, 25 March 2017 (UTC)

  • I'm not sure how someone who's been following this talk page could fail to understand what I wanted. But I'll rewrite it for someone totally unfamiliar with the dispute:
  • Recently, about ten editors active on WT:V had a discussion ending in this diff, as a result of which they made this edit. The effect of this was they changed the longstanding phrase "Verifiability is not a guarantee of inclusion" to "Verifiability is not a reason for inclusion"; with the obvious implication that removing sourced content from established articles is permissible. Was this a good idea? Please support position #1 or position #2 below, or add your own alternative.
  • Position 1: Verifiability is not a reason for inclusion (status quo based on recent talk page discussion). If this position enjoys consensus then the outcome would be to leave the policy as it is.
    *Position 2: Verifiability is not a guarantee of inclusion (longstanding wording). If this position enjoys consensus at the RfC then the policy would be rolled back to this revision.
  • Alternatives/threaded discussion.
  • If this is still unclear then I suppose I could, err, type it slower. ;)—S Marshall T/C 23:47, 25 March 2017 (UTC)
You should remove the non-neutral wording. BrightRoundCircle (talk) 23:52, 25 March 2017 (UTC)
What according to you is non-neutral?—S Marshall T/C 23:55, 25 March 2017 (UTC)

Definite objections. As presently formulated (above), this would be one of the worst RfCs I have seen, with no clear statement of the purpose, or even what these postions are about, or what shoould (or should not) ensue. Marshall's "positions" don't even suggest (for the benefit of editors that coming here cold) that what he wants is to revert a change for which there was discussion and consensus; there is no neutral summary as a basis for the discussion. Nor is there much liklihood of useful discussion, as Marshall still has not responded to the questions presented to him, has not defined his terms (such as the difference between "guarantee" and "virtual guarantee", nor has presented any instances of any relevant edits. As to what arguments might persuade him, he insists on an absurd requirement (described above) that is effectively impossible. All of this before we even reach the core issue of not a reason for guarantee of inclusion. If we must have an RfC, it should be formulated by someone other than S. Marshall. ~ J. Johnson (JJ) (talk) 23:37, 25 March 2017 (UTC)

  • You don't decide whether I'm allowed to start an RfC or not.—S Marshall T/C 23:47, 25 March 2017 (UTC)
He's not saying you're not allowed; he's saying your attitude towards building consensus has been non-cooperative. BrightRoundCircle (talk) 23:54, 25 March 2017 (UTC)       Yes. ~ J. Johnson (JJ) (talk)
  • Oh, yes, I decided not to get sidetracked into a semi-relevant discussion and didn't answer some questions. I'm probably going to Wiki-hell for it.—S Marshall T/C 23:57, 25 March 2017 (UTC)
Stuart, you certainly may start an RfC. But I strongly suggest that you don't, for the reasons given, plus one more: it's unlikely to get what you want. As BRC said, your attitude towards building consensus has been non-cooperative. You make absurd demands on others, but refuse to respond to very basic questions the rest of us have; discussion fails to progress. In our recent discussion we reached consensus in about as many words we have had here, but here you are not even to the starting line. An RfC would just amplify the discord, and would be disruptive to the encyclopedia. ~ J. Johnson (JJ) (talk) 08:49, 26 March 2017 (UTC)

I might try to be a bit of an observer here. As I recall, I've worked with S Marshall, as I recall, I think that the two of us were the most-bloodied pointy end of the spear in the gigantic effort it took to get rid of "not truth". IMHO they are an direct, blunt action-oriented person towards putting no BS accurate information and descriptions into Wikipedia. But we are on opposite sides on this. IMHO (at least from the examples given) they are overly focused on instances where the current wording might get mis-used to remove assessment-type wording like "pseudoscience" of pseudo-science type things. In this area (setting aside the other arguments) IMHO the older wording caused far far more problems than the ones of the type that S Marchall is focusing on. I may have that all screwed up but that's my take and opinion. Sincerely, North8000 (talk) 01:03, 26 March 2017 (UTC)

Not screwed up; I think you're actually close to the mark. Marshal is indeed "overly focused on instances", and particularly on some very limited set of instances where he objects to the removal of certain material, where he can't come up with any other reason to object. He is so focused on this particular (but so far unspecified) battle that he simply fails to see the flip side, where we can't remove errant bullshit because, hey, it's verifiable! He sees only today's battle, with no awareness that tomorrow's battle may be the other way around. ~ J. Johnson (JJ) (talk) 08:53, 26 March 2017 (UTC)
  • A few days ago, S Marshall asked me to give an actual an example of an edit that introduces information that's verifiable by means of an inline citation to a proper WP:RS, when that information doesn't belong anywhere on the encyclopaedia. Off Wiki life prevented me from responding immediately, but I offer the following edits that were made to our article on the Cutty Sark. The edits were made the day that a large fire damaged the ship (21 May, 2007)... as events unfolded, people kept adding minute by minute updates... including the exact times that things occurred (for example: this edit noting that "a representative of the fire brigade said @ 7:09am that the fire was well under control and that damage was extensive but until the experts can view its unknown just how much has been lost." or this edit, noting that "Aerial video footage @ 7:22am shows extensive damage but seems to indicate that whilst damage is extensive the ship as a whole has not been destroyed in its entirety.
Did our readers really need to know the exact time for each of these occurrences? Perhaps it helped while the fire was unfolding, but was it important to include in the long term? Obviously not. Within a few days, the article was re-written with a more proper historical tone... the comment by the member of the fire brigade was removed entirely, and the bit about the aerial footage showing damage was re-written to omit the time the footage was shot. I hope everyone here would agree that the re-writing was appropriate.
My Point?... the fact that the fireman made his comment at a specific time of day is verifiable, and the time of day that the aerial footage was shot is also verifiable. And yet both times were removed. So what justified the removal of this verifiable information? My answer is: "editorial discretion". The exact times were no longer needed once the article shifted focus from reporting on an on-going event to giving an account of a past event. It was appropriate to omit this verifiable information... even though it is verifiable. Furthermore, I can not think of any other article where such trivial information should be mentioned... even though it is verifiable. Blueboar (talk) 13:16, 26 March 2017 (UTC)
Agree, all trivia, all NOT material, all material excluded by BLP, etc, etc. -- Alanscottwalker (talk) 14:16, 26 March 2017 (UTC)
Not as clean or clear of an example as could be desired, but it's a start. (And thanks to Blueboar for presenting it.) It presents two key elements of the matter: material ("information") that is 1) verifiable, and 2) generally held to not "belong" in the article and therefore should be excludable. But it does not appear from the history that there was any issue contingent on verifiability. Even if there had been, the problem with this example is that there are bases for exclusion, such as WP:NOTNEWS, so it isn't a case of verifiability being the sole basis for retaining it. As such it is still a useful example in showing that, at a given location, at least some verifiable content should not be guaranteed inclusion.
I think the kinds of examples we need are where "verifiable" (or perhaps "WP:V") are found in the edit summary, and perhaps someone with more skill than I could search some article histories for such. Marshall would greatly facilitate this discussion if he would point to some actual article edits (not your own rambling commentary, Stuart) where this issue came up. ~ J. Johnson (JJ) (talk) 21:33, 26 March 2017 (UTC)
You really should have had an RfC at the time, for a substantial edit to a major policy of that kind. I'd like you to stop trying to set preconditions for having that RfC now, please.
Blueboar's example is interesting and useful, and it's the kind of thing I was hoping would be brought up here because it might finally lead to a productive discussion. I haven't been able to find the cite for that exact fact --- to be fair, we're dealing with an edit that's nearly ten years old --- but I've found articles like this, this and this from the BBC, this from The Telegraph, and this from The Guardian. I recall the 2007 Cutty Sark fire well --- I was employed in London at the time and the incident was within a mile of my place of work. It was an event of national importance. I agree that the fact cited doesn't belong in Cutty Sark. But I could certainly write an article called 2007 Cutty Sark fire based on reliable sources, and a detailed timeline would not be out of place there. If information comes from a fully reliable source within the meaning of WP:RS, then there is almost always a place for it on Wikipedia. Even if that place is currently a redlink.—S Marshall T/C 23:50, 26 March 2017 (UTC)
The examples Blueboar provided might indeed be well included in the (so far hypothetical) article 2007 Cutty Sark fire. Even better, for the sake of argument we might hypothesize some situation where it is notable that one of the firemen is Roman Catholic. But none of that should guarantee the inclusion of such details in the Cutty Sark article, or any other article. That such details are verifiable, and may even "belong" (however that is defined) in some other article, is no basis for including them in Cutty Sark. If the place for such information is "currently a redlink", it means precisely: that place does not exist. The proper place for information for future or hypothetical articles is on your own computer, not squirreled away in some other article where it has no reason for inclusion. ~ J. Johnson (JJ) (talk) 09:08, 27 March 2017 (UTC)
  • Remember that the edit I'm trying to make is to restore the phrase: "Verifiability is not a guarantee of inclusion". When you answer my view with phrases such as none of that should guarantee the inclusion of such details, you're attacking a position I do not occupy. It's true that I have, somewhat tangentially to the main thrust of my argument, remarked that content that's genuinely verifiable to a fully reliable source almost always belongs somewhere on Wikipedia (even if that place is a redlink at the moment). I have not said that all sourced content belongs in an existing article.—S Marshall T/C 12:11, 27 March 2017 (UTC)
Whatever position you "occupy" seems rather myopic, but it's hard tell because of your refusal to clarify it. E.g., your initial statement in this thread (01:47, 23 Mar) was: "If a thing is verifiable (by way of an inline citation to a reliable, published source) then it probably belongs somewhere in the encyclopaedia.. Which you then followed up with "verifiability should be a virtual guarantee of inclusion", with no qualifications other than RS (which we assume is inherent in V). Only after you have further stated that "(2) Everything verifiable should be somewhere in the encyclopaedia" do you allow the provisions of appropriate placement, due weight, and mainstream view. While your denial is true, it is also irrelevant, because heretofore you have not said anything about existing articles. (In passing I note the additional new qualifiers of "genuinely verifiable" and "almost always belongs somewhere".) While you would have the text read "Verifiability is not a guarantee of inclusion", it is a fair read of your statements that, for a large class of possible content, you would guarantee inclusion. Such inconsistent, unclear, and tangential statements do not advance your argument. ~ J. Johnson (JJ) (talk) 22:17, 27 March 2017 (UTC)
If you don't understand my position then what grounds do you have for calling it "myopic"? What I said right at the start was:- I feel that yes, verifiability should be a virtual guarantee of inclusion, provided the source is properly reliable. You took this to mean "inclusion in a particular article" and, as I made extremely clear in a number of subsequent edits, I meant by it "inclusion somewhere in the encyclopaedia". This is a perfectly coherent, consistent and intelligible view, although you seem peculiarly outraged by it. It's also tangential to the main thrust of my argument because the edit I want to make is to restore the phrase "verifiability is not a guarantee of inclusion", because I do agree and have always agreed that it's permissible to remove sourced content from a particular place.
I suggest that we can close this whole line of argument off, though, by agreeing that the final text version should include a qualifier along the lines of either (1) "verifiability is not a reason for inclusion in any particular article" or, preferably, (2) "verifiability is not a guarantee of inclusion in any particular article".
I agree that it will be helpful if I produce examples that explain what I'm saying. I expect that you will want to interrogate and dispute them so I'll start a separate heading for that shortly.—S Marshall T/C 09:48, 28 March 2017 (UTC)
I believe we all understand and accept the qualification that verifiability in a reliable source is an overall requirement, that we are not talking about material that is unsourced, nor sourced only in non-reliable sources. Your "provided the source is properly reliable" is not the issue (and it is irksome to have to keep specifying that). The issue is your insistence on a guarantee (or the as yet not distinguished virtual guarantee) of inclusion ("belongs somewhere") for RS material not otherwise excluded.
I have been quite aware that what you mean is "inclusion somewhere"; that "somewhere" is part of the issue here. Quite aside for your self-admiration of being "perfectly coherent", it seems you have failed to grasp that inclusion of material is always in local context, in support of specific content. I.e., having found something "interesting" in some reliable source, we do not (well, should not) proceed to find some place to stick it into the encyclopedia. There is no "right of inclusion", whether globally or locally.
The issue here is NOT that "it's permissible to remove sourced content from a particular place", but the "gift" you object to giving "editors who're advancing the best interests of their nationality, religion, preferred conspiracy theory or favourite brand of snake oil (delete as appropriate) and would like to remove information that disfavours it." (20:24, 23 Mar; see also 11:24, 24 March) of removing content that HAS NO REASON FOR INCLUSION BEYOND VERIFIABILITY. Unsourced content, content from unreliable sources, and even content from reliable sources that has no other reason for being included at that spot should be removable by any editor, regardless of whatever agenda they may have. That is the essence of the issue, and what you have objected to. ~ J. Johnson (JJ) (talk) 22:23, 28 March 2017 (UTC)

Compromises?

First my take on the two choices, neither language is "perfect", and at least to my mind, they are basically the same, although I do think one is somewhat clearer (I think it's especially clearer in light of a further sentence that has long been in this policy that no-one disputes - that content has got to meet multiple policies and considerations - I can go at length about why this policy stresses that, if anyone wants, but I am sure all experienced editors know why). Before we get into binary thinking this or that, let's see if there are ways to compromise. SMarshall, as you seem worried about "what to do" when there is removal of a RS'd entry, is there some language your would like in this policy to actually address that (perhaps in a section entitled - removing verifiable material)? Or if you cannot abide "reason", is there a synonym, or adjective, or modifier that would help? Alanscottwalker (talk) 13:59, 26 March 2017 (UTC)

  • I'd like it to encourage and empower editors to add reliably-sourced information to Wikipedia articles, to a greater extent than "verifiability is not a reason for inclusion" does. I'd like it to be clearer that editors should get consensus before removing reliably-sourced information from established articles, and that it's not okay to use this clause to remove reliably-sourced information about climate change, or to sanitise Wikipedia's article about your political party, or otherwise to advance a position. As a secondary point, I'd also like to reverse the part of this edit which unwisely changed "information" to "content". "Information" is the correct word. We don't want people removing photographs or media files because they aren't cited.—S Marshall T/C 00:00, 27 March 2017 (UTC)
See, if you'd have done away with the all-or-nothing approach, you could have achieved consensus already. The change from "information" to "content" is a minor one and the consensus seems to be that "information" is better, so at least in that one regard you'd have already gotten what you want. As for your your first part, the entire rest of the policy encourages adding verifiable information. You should forget about "established articles"; either there's consensus or there isn't. If someone removes verifiable information against consensus, it should be restored until a new consensus is reached. If someone adds verifiable information against consensus, it should be removed until a new consensus is reached. There is no policy for "established articles", only consensus. BrightRoundCircle (talk) 10:49, 27 March 2017 (UTC)
S Marshall, with your "higher bar for removal" one thing to keep in mind is that material can material can be used to mis-inform or obscure information. Let me revers your one example to illustrate. Remember that wp:RS criteria do not specify actual reliability, at best it is ham-handed, defining it with nor regard to the material which is citing it. So, in you example if someone found a wp:rs that said the intelligent design is a science you would have to meet the very high bar of a consensus in order to remove it. I think that you are getting, but they arise from shortcomings in other policies and guidelines plus one or two missing guidelines. WP:VER merely establishes one criteria for inclusion and it can't fix all of those things. North8000 (talk) 11:36, 27 March 2017 (UTC)
I'm sorry, I'm incredibly confused by this. "WP:RS criteria do not specify actual reliability"? What do you think they specify?—S Marshall T/C 12:15, 27 March 2017 (UTC)
They specify the rulebook to be considered to be a "wp:rs" which can end up with a actually reliable or unreliable source. Most key requirements are that it be secondary, published, with a layer of editorial review. So if the food editor in in old fashioned newspaper wrote that Intelligent Design is a science, in Wikipedia that would be considered to be an official WP:Reliable Source. There is no criteria for objectivity or expertise regarding the items which cited it. North8000 (talk) 12:45, 27 March 2017 (UTC)
And even at RS's most limited, what are the chances that 'Food Editor said, he 'believed Intelligent Design'" should go anywhere on the Pedia? Alanscottwalker (talk) 14:34, 27 March 2017 (UTC)
Really? You genuinely don't think WP:RS covers that kind of situation? I'm not sure how it could be more clear.—S Marshall T/C 16:33, 27 March 2017 (UTC)
Don't cover what situation? Basically anything ever published is RS for its author's (or subject's) sometimes trivial, just useless, or false thoughts, but that means it is most unlikely many (most) of those thoughts will find a place in any encyclopedia article, even though they are unmistakably WP:Verifiable. Alanscottwalker (talk) 16:47, 27 March 2017 (UTC)
There is no criteria for objectivity or expertise regarding the items which cited it. — sure there is. WP:SCHOLARSHIP, for example, lets you argue that the opinion of the food editor in a non-scholarly newspaper is not an appropriate source (even if it's technically a "reliable source") for a scientific matter. If you don't like guidelines and want a policy, there's WP:EXCEPTIONAL and WP:ONUS and WP:VALID and WP:BESTSOURCES. If any piece of verifiable information were put on Wikipedia it would be a very shoddy encyclopedia. BrightRoundCircle (talk) 16:53, 27 March 2017 (UTC)
I'd love to continue on this but I think that this is a tangent except that I raised it to show an unintended consequence of a proposal. If I continued, I'd ask to to show me something in policy strong enough to knock out my sourcing example if I fought to keep it in.  :-) Sincerely, North8000 (talk) 17:10, 27 March 2017 (UTC)
If you did, then I'd say that WP:RS covers it (at WP:NEWSORG, specifically). Since you don't wish to continue, may I take it that you concede my point?—S Marshall T/C 18:16, 27 March 2017 (UTC)
Nope, I'd say to point out the specific place that excludes my source. I'm not talking about general good advice, I'm talking about something to stop "my" wiki-lawyering to keep it in.North8000 (talk) 19:53, 27 March 2017 (UTC)
You won't find something that stops wiki-lawyering. In all but the most trivial cases, applying any Wikipedia policy about sourcing and verifiability appropriately means judgment calls, and some idiot will be able to lawyer any guidance like that. Don't write policy for extreme cases involving people with an agenda; hard cases make bad law. The purpose of policy is to educate good faith editors about good practice and to help them think appropriately about attribution and sources. Those editors should be encouraged to add sourced content to articles. Telling them "verifiability is not a reason for inclusion" is inappropriate.—S Marshall T/C 20:04, 27 March 2017 (UTC)
Then perhaps you would provide some easy cases, actual article edits where an issue of adding or removing content hinged solely on verifiability. By the way, given this copious evidence of voiced disagreement, do not presume consensus from future silence. The disagreement is established, we are not required to continually renew it. ~ J. Johnson (JJ) (talk) 22:40, 27 March 2017 (UTC)

It's never going to be solely about verifiability with a real world edit, because editors always introduce other considerations during the discussion. You'll need to accept examples where WP:V was a major limb of the dispute, but there are also other policies and guidelines applicable. Agreed?—S Marshall T/C 22:45, 27 March 2017 (UTC)

No, I don't agree... some bit of information can be verifiable, and yet not considered important enough to the article topic to be worth mentioning.
For example, historians might know what King Richard III ate for breakfast the morning before the Battle of Bosworth Field... but is this something we think is worth mentioning in our article on the battle? (I would say no). That sort of removal is not a policy or guideline based determination, but a matter of editorial judgement. Blueboar (talk) 23:49, 27 March 2017 (UTC)
Yes, but the key words there are "the article topic". The information if sourced to a genuinely reliable source, would belong somewhere else.
To be fair, it's incredibly unlikely that a historian would find evidence of Richard III's last breakfast, but okay, let's stipulate for the sake of argument that someone did. The information would certainly be published in reliable sources, as well as the popular press. Contrary to what you say, it might actually be suitable for inclusion in Richard III, if it was remarkable enough. Our article on Henry I does devote the best part of a paragraph to what was presumably his last meal. (Famously, Henry I died of a surfeit of lampreys.)
But okay, let's stretch our imaginations even further and say that we can source Richard III's last breakfast and it was toast. Yes; at that point we have a factoid that would not belong in Richard III. It would belong in History of breakfast.—S Marshall T/C 00:12, 28 March 2017 (UTC)
I would not include it in History of breakfast... as that article focuses on the typical breakfast through the ages (and Richard's may not have been typical... even for a King). Blueboar (talk) 02:05, 28 March 2017 (UTC)
Or, indeed, in Toast#History.—S Marshall T/C 10:04, 28 March 2017 (UTC)
On what basis can you possibly claim it matters at all to the history of toast - if you say it's because you have RS that mentions Henry and toast, you have not answered the question, and rather are suggesting a POV, OR, or trivia push. Alanscottwalker (talk) 12:55, 28 March 2017 (UTC)
The premise of the question is that we have a reliable source of information from 1485. It would be the first mention of toast as a breakfast dish, the first reliable description of a king's breakfast, and the first post-Roman description of what people ate before a battle, among rather a lot else.—S Marshall T/C 13:38, 28 March 2017 (UTC)
You just made all that up about the importance of Henry and toast. That's either OR pushing, POV pushing, or Trivia pushing and belongs nowhere on the encyclopedia. Alanscottwalker (talk) 14:14, 28 March 2017 (UTC)
Oh, I've just seen this. No, I didn't make it up; yes, I do know what I'm talking about; and you're clearly just another J Johnson, disagreeing because it's me saying it.—S Marshall T/C 16:51, 1 April 2017 (UTC)
I'm sorry, where in the literature of toast or breakfast can you produce anything about the importance of Henry? You know what we call edits where people put together such stuff? (Also, I did not say 'you do not know what you are taking about, here' I said, such would be OR pushing, POV pushing, or Trivia pushing) For example, primary sources are verifiable for their contents, but Wikipedia editors don't insert in articles their own 'contributions to the literature' based on such sources, even though they are verifiable. Alanscottwalker (talk) 18:29, 1 April 2017 (UTC)
Stuart: to the extent "real world edits" have other considerations for inclusion you simply do not need verifiability to get something included. Where other reasons or considerations for including something are lacking, it should not be included merely because it is verifiable. Whether some material might "belong somewhere else" is entirely irrelevant, as inclusion is ALWAYS in the context of a specific article. That certain material is suitable for some article in no way guarantees it a place in any other article. ~ J. Johnson (JJ) (talk) 08:08, 28 March 2017 (UTC)

S Marshall, the current wording essentially says that once verifiability has been met, that this particular policy is neutral on inclusion. So all of the other mechanisms would determine whether or not to include something. What's wrong with that? And that is in fact what the body of this policy says. You are arguing to put something into the lede which implies otherwise. North8000 (talk) 10:47, 28 March 2017 (UTC)

  • Yes, and Wikipedia's other policies aren't neutral on inclusion, and shouldn't be. If the material is appropriate in context, then there's a clear presumption to keep it. Arbcom have quite rightly imposed sanctions on editors for removing verifiable material from the encyclopaedia. I'll shortly produce a number of real world examples of why we do keep information that's appropriate in context, and why this wording that makes it easier to remove such material is problematic.—S Marshall T/C 11:23, 28 March 2017 (UTC)
Let me give another example of where we might remove verifiable information from one article, and not find any other article to put it in: As is noted in our article on the Eye of Providence (the symbol depicting an eye surrounded by a triangle, as seen atop a pyramid on the back of the dollar bill ... and often associated in pop culture with the Illuminati), the symbol appears in numerous video games. Now, it could be argued that the article would benefit from some examples of video games that depict the eye. Fair enough... but that raises the question of which video games to mention as examples. There is no reason to clutter up the article by listing every single video game that depicts the symbol. The idea is to give a few, well chosen, examples... not an exhaustive list. And so, we might well decide to mention the more prominent games that depict the symbol, and omit the more obscure ones. That is an editorial decision... not something based on policy or guidelines.
Now... do we need to find a home in which to mention the omitted depictions? No. Unless the symbol is important to the game play of the particular game (and it may not be) the depiction probably would not rate being mentioned in the article on the game itself... Indeed, there may not be a game article in which to mention them... the game may not pass our Notability guidelines for video games.
So... no... verifiability does not guarantee inclusion. Our editors might well reach a consensus to omit (ie not mention) a specific fact... even when that fact is verifiable. Blueboar (talk) 14:16, 28 March 2017 (UTC)
No indeed. I've always agreed that verifiability does not guarantee inclusion, and this is why the edit I want to make is to restore the policy wording that says "verifiability does not guarantee inclusion". This S Marshall who wants verifiability to be a complete guarantee of inclusion sounds like a very silly person; I'm glad I'm not like him!
I have said that information sourced to a genuinely reliable source as defined in WP:RS almost always belongs somewhere on Wikipedia. The qualifiers are important.—S Marshall T/C 15:35, 28 March 2017 (UTC)
S Marshall at 01:47, 23 March: "verifiability should be a virtual guarantee of inclusion". Yes, very silly. ~ J. Johnson (JJ) (talk) 22:32, 28 March 2017 (UTC)

On real-world examples of verifiability disputes

The examples we've discussed so far have been a bit trivial and hypothetical, and I'd prefer to focus on real disputes. In this section I will place examples for discussion.—S Marshall T/C 14:27, 28 March 2017 (UTC)

Purpose of the section

My concern here is about a recent change that turned "verifiability is not a guarantee of inclusion" into "verifiability is not a reason for inclusion". The effect of this change was to make it easier to remove verifiable information from articles, and I think that was probably its intent. I object to this edit and would like it to be reverted. Let's focus on that.—S Marshall T/C 14:27, 28 March 2017 (UTC)

Real world examples are messy

Real world examples aren't solely about WP:V. Editors who're being questioned typically give several justifications for their actions. If an editor removes content under WP:BURDEN, they will often say they also think the content is OR or violates NOT. If they remove content from a biography, they will cite BLP as a reason. I'm going to deal with disputes that involve WP:V as a primary part of what happened, but not as the sole locus of dispute.—S Marshall T/C 14:27, 28 March 2017 (UTC)

Where content is removed for other justifications verifiability is not in issue. The issue is whether material can be retained where there are no other justifications than verifiability. Any "messy" example is not relevant to the issue. ~ J. Johnson (JJ) (talk) 22:57, 28 March 2017 (UTC)
Editing hastily

I think it's better to deal with non-trivial examples, and these often involve people editing hastily, or while overwrought, or in some cases, in anger. Let's try to look past these editors' behaviour, even if that's sometimes difficult, and examine the underlying policy issues.—S Marshall T/C 14:27, 28 March 2017 (UTC)

Example #1

List of awards and nominations received by Philip Seymour Hoffman.

On 21st June 2015, an editor removed, over the course of three minutes or so, 19,485 bytes of unsourced information about his awards. The information in question was verifiable, as evidenced by the fact that another editor went through and verified it; but it was not actually sourced at the time of the removal. There followed a rather unfortunate edit war.

During the subsequent Arbcom case, the removing editor justified his actions in the following terms:- The main issue here is not WP:BLP, however: it's WP:BURDEN and WP:V. The requirement for inline citations after material has been challenged is absolute: it harbors no exceptions. Once material has been challenged on sourcing and validity grounds, it cannot be restored by any editor without an inline citation that supports it. In other words, this dispute was fundamentally about removing verifiable content from an article.

The removing editor was desysopped after the Arbcom case.—S Marshall T/C 14:42, 28 March 2017 (UTC)

An unfortunate situation, which could have been easily avoided ... rather than engaging in an edit war, the editors who wanted to return the removed information could have provided a source when they did so. That would have resolved the issue as far as WP:BURDEN was concerned. The situation might not have escalated to the point where Arbcom needed to get involved. Blueboar (talk) 15:13, 28 March 2017 (UTC)
Regardless of what the editor later said, what people actually determined they saw was not a dispute about verifiability (because there was agreement later that it was all verifiable), it was a dispute about in-line citation -- in-line citation is in fact more a WP:BLP issue in removal because of the express language of the BLP intro. As for Burden, you are already in a different part of policy than is being discussed, so the example is not relevant. If you want to make this relevant, to what is being discussed, you would actually have to change the situation to something like the Hoffman list had a long section of Cary Grant awards. Alanscottwalker (talk) 15:29, 28 March 2017 (UTC)
I think that the discussion we're having is about the relationship between verifiability and inclusion in WP:V. The example is about deciding whether to include verifiable content in an article. Surely it's immensely relevant and totally on point.—S Marshall T/C 15:41, 28 March 2017 (UTC)
The example is not relevant, nor on point: it's an in-line citation dispute, that is all. Alanscottwalker (talk) 15:45, 28 March 2017 (UTC)
I agree with Alan... this wasn't a "Verifiability does not guarantee inclusion" issue. At no point did anyone argue that "It does not matter whether the information is verifiable... this article should not mention it, even if it is verifiable." Blueboar (talk) 16:10, 28 March 2017 (UTC)
Yes! Thank you for that. That's the key point from this example.  :) Everyone, including the removing editor who is the poster child for an extreme WP:V literalist, assumes throughout more than 100,000 bytes of discussion, that if it can be verified, then it should be included. Were they wrong?—S Marshall T/C 18:05, 28 March 2017 (UTC)
Wrong. What was assumed was that the topic was notable because it was not a deletion dispute. But no one argued (or assumed) you could add anything you want to the list -- like say, Cary Grant's awards or Hoffman's list of pets -- merely because they are verifiable. -- Alanscottwalker (talk) 22:20, 28 March 2017 (UTC)
... and that's why I want to restore the policy wording that says "verifiability is not a guarantee of inclusion".—S Marshall T/C 23:50, 28 March 2017 (UTC)
Now that you have shifted to "almost always", we are much closer to agreement than we were. I get the distinct impression that we have not been discussing the same thing. Perhaps we have been talking at (or past) each other instead of actually talking to each other. I think the disconnect is that you have been broadly focused ... on how the current wording might be misinterpreted. while I have been focused narrowly ... on how it is supposed to actually be interpreted. Blueboar (talk) 02:30, 29 March 2017 (UTC)
S Marshall, I am in 100% agreement with your general view on that particular situation, and that such is a problem in Wikipedia. I think that a few evolutions in WP:Ver over the last couple years are starting to help that a bit, but a bit more is needed. However, IMHO it has 0% to do with the current debate about the sentence. This was a case where the remover had the "letter of the law" of wp:ver explicitly behind them and was mis-using that. Sincerely, North8000 (talk) 10:49, 29 March 2017 (UTC)

Well. Was he misusing WP:V? Arbcom clearly felt he was misusing WP:BLP. Whether he was misusing V is less clear to me. This comes bundled with all the stuff about overusing WP:BURDEN target one editor or one article that we've discussed several times before. As you'll probably recall, I've long felt there should be reasonable limits on overusing WP:BURDEN to target people and you two disagree, which is why you've rejected my proposals to put a qualifier on WP:BURDEN.

But even without WP:BURDEN, the way the policy is currently written, people can remove verifiable material from the encyclopaedia at will. Because you've changed "verifiability is not a guarantee of inclusion" to "verifiability is not a reason for inclusion". It's a phrasing that will be misunderstood and applied in ways you did not intend.—S Marshall T/C 11:10, 29 March 2017 (UTC)

Example #2

Child grooming

On 1 April 2016, an editor removed the words Child grooming is an activity done to gain the child's trust as well as the trust of those responsible for the child's well-being. Additionally, a trusting relationship with the family means the child's parents are less likely to believe potential accusations with the edit summary Seems like a lot of original research.

(This still makes me quite upset. In my personal opinion, this was one of the most reckless edits I've ever seen on Wikipedia with considerable real world implications. Because vulnerable children do read Wikipedia, and they tend to overtrust it. I think there are good reasons to be worried about any child who's looking at child grooming, and such a child must be told the truth, no matter what.)

In practice, there was a brief edit war between an editor who was outraged at this removal, and the removing editor who was outraged that it was restored. This then spilled over onto this page, WT:V, where the removing editor justified himself with the words Losing content is the lesser of two evils, when compared to have POV/fringe/incorrect content. If content is removed due to the lack of a source, it can always be restored when a source is located. In other words, he considers it acceptable to remove unsourced material even if that material is subsequently shown to be correct.

The disputed sentence now reads To establish a good relationship with a child and the child's family, child groomers might ... try to gain the child's or parents' trust by befriending them, with the goal of easy access to the child.S Marshall T/C 15:01, 28 March 2017 (UTC)

I remember this one... and that my reaction was that the sentence in question was indeed the type of statement that needed a source. I had no idea whether the information was verifiable or not, but I supported removal, and a return with a source. The debate seemed to be about the accuracy of the statement. Debates over accuracy are always tricky. We do want our information to be both accurate and verifiable... and sometimes that is difficult to achieve. I have found that disputes of this nature can usually be resolved by rewriting the disputed information and phrasing things in a different way (as ultimately happened in this example). Blueboar (talk) 15:30, 28 March 2017 (UTC)
First, if I may try to diffuse the emotion, given the first sentence of the article, the emotional response seemingly makes little sense. And again, insisting on sources in in-line citation is not on point to what has been the language under discussion. Alanscottwalker (talk) 15:39, 28 March 2017 (UTC)
I think that the discussion we're having is about the relationship between verifiability and inclusion in WP:V. The example is about deciding whether to include verifiable content in an article, isn't it? To my mind this is absolutely germane. How is it not?—S Marshall T/C 15:44, 28 March 2017 (UTC)
Your first sentence makes no sense, perhaps you mistyped. The discussion is about, whether just because a piece of content like, say, 'The cat got out of the tree.' can be verified (with no doubt multiple RS) that is the reason it should go into any article. Alanscottwalker (talk) 16:06, 28 March 2017 (UTC)
Up there I wrote: Purpose of the section: My concern here is about a recent change that turned "verifiability is not a guarantee of inclusion" into "verifiability is not a reason for inclusion". The effect of this change was to make it easier to remove verifiable information from articles, and I think that was probably its intent. I object to this edit and would like it to be reverted. Let's focus on that. I assure you categorically that this is the subject about which I'm talking.  :) Are you on a different topic?—S Marshall T/C 18:08, 28 March 2017 (UTC)
S Marshall, if I were there I would have worked very hard to get the grooming sentence back in by sourcing it. But either way, IMHO it has 0% to do with the current debate. The remover had pretty well the full weight of WP:ver behind them, except one phrase that IMHO had absolutely nothing to do with it is the the one that we are currently debating. North8000 (talk) 10:58, 29 March 2017 (UTC)
S Marshall, whatever you want to talk about you are not bringing forth examples that deal with the language under discussion. In this example, if you can separate your emotion, you will see that someone removed something on 'original research' grounds -- the text was then apparently improved and clarified, cited, and revised (so what's to complain- we are here to improve the encyclopedia). And, at any rate, it was an Original Research issue, not the sentence in this policy. Alanscottwalker (talk) 14:20, 29 March 2017 (UTC)

Example #3?

Example 3 is a placeholder for the time being. The above two are not the only examples I have, but I will stop here because I'm conscious that I'm swamping the whole talk page with posts by me. I would like to reserve this spot for future examples if they become necessary over the course of the discussion.—S Marshall T/C 15:01, 28 March 2017 (UTC)

How about an edit where someone "advancing the best interests of their nationality, religion, preferred conspiracy theory or favourite brand of snake oil" removes "information that disfavours it"? That is, after all, your original concern. Or how about removal of "psuedoscientific" from any article, where verifiability was the sole reason for retaining it? ~ J. Johnson (JJ) (talk) 22:59, 28 March 2017 (UTC)

Sure. On 1st December 2014, an editor removed the phrase "Major Pakistani Victory" from the infobox in Battle of Chawinda and replaced it with "Stalemate". Interestingly, he left the citation for "Major Pakistani Victory" in place when he did this, but added another source of his own choosing as well. An unfortunate edit war followed, as is so often the case with real world examples, followed by a talk page exchange in which the removing editor justified his removal of information that disfavoured India on the grounds that the source presented was unreliable. This is, at bottom, a verifiability dispute, although the editor -- who over the course of his two year Wikipedia career became a rollbacker with nearly 200,000 edits, and is now indefinitely blocked for socking -- also mentioned OR as part of the dispute.—S Marshall T/C 23:47, 28 March 2017 (UTC)

Bullshit. That little edit tiff was about what the source said (with edit summaries like "Read the source" and "Read the source yourself"), or whether the source was reliable. Deleting material because is not reliably verified (which we agree is allowed) is not the same as deleting reliably verified material because there are no other considerations for retaining it. You have yet to present any examples of the latter. ~ J. Johnson (JJ) (talk) 00:48, 29 March 2017 (UTC)
This illustrates a common problem - wiki-lawyering to POV an article. And IMO some policy and guideline changes are needed to reduce that problem. But again, the wikilawyer used other areas of wp:ver, wp:nor and wp:rs to further their actions. IMHO it has 0% to do with the phrase that we are debating. Sincerely, North8000 (talk) 11:06, 29 March 2017 (UTC)

Example #4?

Example 4 is a placeholder for the time being. The above three are not the only examples I have, but I will stop here because I'm conscious that I'm swamping the whole talk page with posts by me. I would like to reserve this spot for future examples if they become necessary over the course of the discussion.—S Marshall T/C 23:47, 28 March 2017 (UTC)

Example #5

I reverted an addition to Year 2000 problem because, although verifiable, it wasn't interesting enough for that article. And although I didn't mention it in my edit summary, I don't think it's interesting enough to mention in any article. Jc3s5h (talk) 02:09, 29 March 2017 (UTC)

Hypothetically, someone reverting your edit could have argued that the material was "interesting enough" for inclusion. Which is a different argument than "it's verifiable, therefore guaranteed inclusion, interesting or not." ~ J. Johnson (JJ) (talk) 19:45, 29 March 2017 (UTC)

Summary on examples

S Marshall I'm mostly with you on those examples, but IMHO they clearly have absolutely nothing to do with the phrase that we are debating. And in fact the types of problems that you are discussing (in essence wiki-lawyering to either POV an article or to conduct a pissing war) are widespread and need some changes to fix them. But IMHO the phrase that you are debating is absolutely not related to that problem. And in fact, your idea would worsen the other 1/2 of the "wiki-lawyering to POV an article or to conduct a pissing war" situations/problem where people want to force in crap.

If you really want to fix the problems that you are discussing, I'd drop the debate on the phrase, and let's go to work on some things that would actually help. One starting point would be to give a current sentence from wp:ver more strength and prominence, because it is always ignored: "When tagging or removing material for lacking an inline citation, please state your concern that it may not be possible to find a published reliable source for the content, and therefore it may not be verifiable". I think that this would actually further your goals unlike (respectfully) your current pursuit of that phrase. North8000 (talk) 11:21, 29 March 2017 (UTC)

I think North has a point. To attempt to generalize the conversation... what S. Marshall appears to be concerned about are situations where the removal of verifiable information was inappropriate. What the rest of us seem to be concerned with are the situations where it is appropriate to remove verifiable information. We are not talking about the same thing, and thus have been talking at each other and not with each other. S Marshall's concerns are valid... however, I don't see how the section under discussion relates to what he is concerned about. Blueboar (talk) 12:29, 29 March 2017 (UTC)
Yes. But see my comment further on. ~ J. Johnson (JJ) (talk)
  • Well, discussing the phrase is gaining no traction with you; no traction with BrightRoundCircle who wants "not a reason for inclusion" to be an unchallengeable fait accompli; no traction with J Johnson whose only interest in the debate now is to antagonise me; and no traction with AlanScottWalker who's provided no diffs of his own and doesn't accept that any of my evidence is relevant. I'm not clear on whether it will gain traction with Blueboar; past experience makes me hopeful that he may be susceptible to evidence. Has yielded one productive example from Jc3s5h which might be worth exploring (he's finally presented evidence of an edit that does justify "verifiability is not a reason for inclusion"!) although I agree with WAID that we shouldn't be changing policy on the basis of that one edit. All in all I'm going to go ahead and say the examples won't end in consensus.
    But now we have a policy that says "verifiability is not a reason for inclusion" and I simply don't think the community agrees. I'll give it a couple of days to see if someone has a killer point to raise, and if not, start an RfC with a view to reversing the change.—S Marshall T/C 12:41, 29 March 2017 (UTC)
Well, as I said, I actually am not wedded to the language (and also had no hand in that revision being put in the policy), but yes your points have been either irrelevant to the sentence or just poorly expressed. When someone objects to something not being cited in-line, you may not like that challenge for whatever reason but it makes no sense to try to argue that challenge cannot be made because of your odd claim that every bit of everything ever published belongs in the pedia somewhere. Similarly, when someone objects that something is original research, again you may not like that challenge for whatever reason, but again it makes no sense to bar that challenge because of your odd theory. Or when someone objects on BLP grounds or NOT or, etc. etc. (again the same). And those objections you now make as examples are obviously not based on the rewritten sentence, or its predecessor. Whatever your objective, what you seem to be arguing again and again is that the pedia cannot be improved by removal - once someone puts whatever scrap of info they want into an article. Alanscottwalker (talk) 13:19, 29 March 2017 (UTC)
  • Are you turning into J Johnson? "Every bit of everything ever published belongs in the pedia somewhere" is a gross distortion of my position. And I'm not trying to ban people from making a challenge, wherever did you get that idea from?—S Marshall T/C 14:24, 29 March 2017 (UTC
No. It's what I understand your position to be, because of what you have argued, and the examples you have brought forward to find fault with - eg. someone removing because of lack of in-line citation, someone removing because of "original research" - if you agree those are useful challenges to improve the encyclopedia, than there is nothing wrong with removal challenge - and your disapproval makes little sense, on top of not being related to the sentence under discussion . Alanscottwalker (talk) 14:44, 29 March 2017 (UTC)
You're right to say that I'm concerned about inappropriate and unwise challenges, and I have provided examples of people removing verifiable content when they should not have done. Why have you leapt from this to saying that I want to ban bar people from making challenges?—S Marshall T/C 15:07, 29 March 2017 (UTC)
Because we are working with humans, here. That means they are sometimes right and sometimes wrong - we can't make them be always perfect (in particular perfect to you, or perfect to me). They can remove, they can add - sometime their additions will be crap (or just not good), sometime their removals will be crap (or just not good) - the balance between those has to even-out. Alanscottwalker (talk) 15:15, 29 March 2017 (UTC)
... and that's why you think I want to ban bar people from making challenges? I'm afraid I'm having trouble matching the answer to the question.—S Marshall T/C 20:25, 29 March 2017 (UTC)
You seem to want to prevent removal. Removal is a challenge. Ergo, you want to prevent challenge. (Also, 'ban' is not a word I have used.) Alanscottwalker (talk) 21:08, 29 March 2017 (UTC)

Bar, then. This is the fallacy of the excluded middle. I want wording that discourages people from recklessly or negligently removing verifiable material from articles, but enables them to do so in a thoughtful and reasoned way. This is why I see "verifiability is not a reason for inclusion" as a step backwards from "verifiability is not a guarantee of inclusion".—S Marshall T/C 21:57, 29 March 2017 (UTC)

Discouraging people from recklessly or negligently adding material would make removal moot. At any rate, I just removed by copy-edit a 30 year old high school award from an academic's biography (regardless if it's verifiable, it's trivial and removal improved the pedia, imo), but feel free to re-add it (and verify it) and we can discuss it, if you like. -- Alanscottwalker (talk) 22:22, 29 March 2017 (UTC)
The policy does discourage people from recklessly or negligently adding unverifiable material. It says "Even if you're sure something is true, it must be verifiable before you can add it" (which is phrasing that encapsulates the necessary concept beautifully, if I do say so myself). It doesn't discourage people from recklessly or negligently removing verifiable material that's appropriate in context. (The editing policy does. At WP:PRESERVE it says "Preserve appropriate material". There's a tension between this and the implied licence to remove information that WP:V currently contains.)—S Marshall T/C 10:53, 30 March 2017 (UTC)
This policy in WP:PROVEIT, does discourage removal of appropriate content. As for implied licence, the licence is to edit, and it's found in the software, everything else, policy and guideline seeks to restrict that. Alanscottwalker (talk) 11:09, 30 March 2017 (UTC)
But logically, WP:PROVEIT no longer discourages removal of appropriate material. As you rightly say, WP:PROVEIT puts an onus on the restoring editor to demonstrate verifiability --- but right now, we also say that verifiability isn't a reason to include the disputed material! So even if they do provide an inline citation to a reliable source, they still won't get to retain it in the article unless there's a talk page consensus to say they can. We've reached the stage where you need an informal permission slip to add content to the encyclopaedia.—S Marshall T/C 22:38, 30 March 2017 (UTC)
This isn't new and the change in wording of "reason"/"guarantee" didn't bring this about. Consensus works that way regardless of these changes to WP:V. If your bold change gets staunchly reverted, the only way to resolve the dispute is to discuss. Neither version of WP:V negates this, and verifiability has never been a "reason"/"guarantee" to include disputed material. WP:PROVEIT specifically deals with material that had its verifiability challenged, not material that was removed for other reasons. BrightRoundCircle (talk) 00:36, 31 March 2017 (UTC)
Nope, fraid that's not correct. While there are special cases such as WP:BLP or WP:COPYVIO that enable you to remove text on sight, most cases are covered by the editing policy which shows that you should preserve it if it's appropriate; dispute it if it's inappropriate; and fix it if it's inaccurate. Removing it is covered under WP:CANTFIX which says that under WP:V you can remove "unsourced and contentious" material. It also gives other grounds for removing material under other policies, of course, but if the information isn't "unsourced and contentious" then removing it under WP:V would require a talk page consensus in favour of removal. The default is to keep the information in some form.—S Marshall T/C 17:27, 31 March 2017 (UTC)
But since "verifiable" is not a synonym of "appropriate", than it is not appropriate to pretend that they are, as your arguments seem to be doing. Alanscottwalker (talk) 19:00, 31 March 2017 (UTC)
If the information is inappropriate, then you can remove it under WP:NOT. If your concern is simply about inappropriate information, then you're editing the wrong policy.—S Marshall T/C 15:09, 1 April 2017 (UTC)
Actually, you are editing the wrong policy or just confused. Because it is your arguments that keep eliding verifiable and appropriate, as if they are the same thing. (As for NOT, your limiting categorization of it is wrong, but also irrelevant). -- Alanscottwalker (talk) 15:44, 1 April 2017 (UTC)
The majority of well-sourced additions to Wikipedia are appropriate, and if someone does remove well-sourced information, it's a good idea to look at why they've done it, in case they're advancing an agenda that doesn't have anything to do with improving an encyclopaedia. I'm curious. In what sense do you think I've made a limiting categorization of WP:NOT?—S Marshall T/C 16:49, 1 April 2017 (UTC)
I am glad you agreed that when "well-sourced information" (content) is removed, "it's a good idea to look at why they've done it". Which is one reason I have asked for examples where verifiability has been the sole basis put forward for removal or retention. I strongly suspect that in all such cases, where there is strong case for either appropriateness or inappropriateness, other considerations are sufficient (even WP:NOT in some cases though not in others), and it is not necessary to invoke verifiability. Show me any case where removal is clearly appropriate, and I am pretty sure I can show you a basis for removal other than WP:V. ~ J. Johnson (JJ) (talk) 21:42, 1 April 2017 (UTC)
WP:V would require a talk page consensus in favour of removal. The default is to keep the information in some form - that's just not true. "Consensus may determine that certain information does not improve an article, and that it should be omitted or presented instead in a different article. The onus to achieve consensus for inclusion is on those seeking to include disputed content." This is in the version you support. Bright☀ 16:01, 1 April 2017 (UTC)
By the way, this incredibly important sentence ("consensus for inclusion") was never discussed! It was silently accepted without a single discussion before or after it was added. This is the first time the phrase is mentioned on the talk page! Bright☀ 16:20, 1 April 2017 (UTC)
Agree, inclusion is not the default. Quite the opposite. ~ J. Johnson (JJ) (talk) 21:47, 1 April 2017 (UTC)
The only "license to remove" in WP:V is inherent in the requirement that all content (aside from uncontested blue sky or common knowledge stuff) must verifiable in reliable sources. Whether material meeting that minimal requirement should be removed, or added, is not within the scope of WP:V; such matters are left for other policies. If there is no other reason for including/retaining something the exclusion/removal is warranted not on WP:V, but on there being no other reason for retention. (How many times do we have to explain this to you?) Where any content passes the WP:V test any question regarding its removal or inclusion is addressed at other levels. Using WP:V for such purposes only weakens it in its basic purpose.
Whether verifiable material is removed (or added!) "recklessly or negligently", or carefully and deliberately, is absolutely irrelevant here. That is quite unnecessary verbiage.
BTW, your reading of WP:PRESERVE is skewed. That section (of WP:Editing policy) is titled "Try to fix problems", and it starts: "Fix problems if you can, flag or remove them if you can't." Please note that the injunction to preserve material applies to appropriate material. Note also that following section (WP:CANTFIX) points to several "Problems that may justify removal". Which all demonstrates: WP:PRESERVE is NOT absolute. The appropriateness of any material is addressed by other policies. ~ J. Johnson (JJ) (talk) 21:17, 30 March 2017 (UTC)


Blueboar and North8000 have good points, but actually we are talking about the same thing, just different aspects. Marshall suggests several cases where removal seems inappropriate, his concern being that any weakening of WP:V would be a "gift" to those who would do inappropriate removals, even asserting that verifiability itself should guarantee inclusion. However, he has not grasped the flip-side of this, the appropriate removal of inappropriate material. He wants to use WP:V against inappropriate removal, but that is using the wrong tool, and fails to consider the impact on appropriate removal. Or, alternatively stated, inappropriate inclusion. ~ J. Johnson (JJ) (talk) 21:55, 29 March 2017 (UTC)

Recap

For those of you joining us just now, this is a discussion about the difference between these two versions:

Current consensus ("reason")
while verifiability is required for including something, it is not a reason for inclusion.
Verifiability is not a reason for inclusion

While content must be verifiable in order to be included in an article, verifiability alone is not a reason for inclusion, and does not guarantee that any content must be included in an article. Consensus may determine that certain information does not improve an article, and that it should be omitted or presented instead in a different article. The onus to achieve consensus for inclusion is on those seeking to include disputed content.

Previous consensus ("guarantee")
while verifiability is required for including something, it does not guarantee inclusion, which may be influenced by other policies and guidelines.
Verifiability does not guarantee inclusion

While information must be verifiable in order to be included in an article, this does not mean that all verifiable information must be included in an article. Consensus may determine that certain information does not improve an article, and that it should be omitted or presented instead in a different article. The onus to achieve consensus for inclusion is on those seeking to include disputed content.

Please help. BrightRoundCircle (talk) 13:36, 29 March 2017 (UTC)

Sigh. I was kind of planning to keep my mouth shut on this whole...discussion, but I guess now's as good a time as any to offer my two cents. I'm going to number these in case anyone wants to reference them subsequently.
  1. In terms of "bottom line", I see little difference between the two boldfaced phrases, and would like to see an example where the variance would cause an actual problem. Ideally a real example versus a hypothetical one. To me, they're both essentially saying "don't use 'this information can be verified!' as a defense against deletion". As I think about this more, the only consideration that's coming to mind is that "reason" seems to be prescriptive while "guarantee" is more after-the-fact. That is to say, "don't add information just because it's verifiable" versus "don't be surprised that your information got deleted if all you can say about it is that it's verifiable".
  2. I think I prefer "content" over "information", but I could be swayed on that wording.
  3. I prefer "verifiability alone..." over "this does not mean..." though I might argue that the 'alone' is superfluous.
  4. Perhaps instead of 'any content must be included' we might use 'any content is appropriate for inclusion'. Just a thought. OTOH, "does not mean that all verifiable information must be included" strikes me as stating the obvious, and I'd hope no editor ever claimed "this must be included because it can be verified!"
Hope this helps in some manner. DonIago (talk) 15:17, 29 March 2017 (UTC)
The recap looks more like a goal, as so far we have been dealing almost entirely with the overburden, and haven't gotten to the sublime distinctions between "reason" and "guarantee".
As a side note on "content" versus "information": while we hope our content conveys information, that is not necessarily so. "Content" is the more general term, which avoids certain implications of "information". ~ J. Johnson (JJ) (talk)
"Alone" is not merely superfluous, it suggests that verifiability can carry some weight beyond being just the least of what is required. As to your last phrase, unfortunately that is a fair imputation of what Marshall has said and argued. ~ J. Johnson (JJ) (talk) 19:44, 30 March 2017 (UTC)
Yes, and that is a main point. And besides reinforcing a false and harmful urban legend, it puts / implies something in the lede which (rightly so) is not in the actual policy. North8000 (talk) 20:31, 30 March 2017 (UTC)

Let's clear another point of contention...

  • Material ("content"/"information") may be removed if its verifiability is challenged. It can then be restored if it's properly cited.
  • Material may be removed for other reasons, even if it's properly cited. (examples of such reasons can be found in WP:NOT and WP:CONSENSUS, among other policies)

The above changes to WP:V do not change either of these. There appears to be a misconception that these changes to WP:V would somehow make it more difficult to add material to Wikipedia or protect material from being removed. This isn't the case. The changes are only an attempt to make the wording clearer, and they do not affect the verifiability challenge-process or the other-than-verifiability challenge-process, both of which remain exactly the same. BrightRoundCircle (talk) 00:46, 31 March 2017 (UTC)

I very much prefer "content" to "material". Other than that, no news for me here. :) DonIago (talk) 03:40, 31 March 2017 (UTC)
I don't object to "material", but am feeling a slight preference for "content". ~ J. Johnson (JJ) (talk) 08:37, 31 March 2017 (UTC)
I have no preference regarding "material"/"content"/"information", I just tried not to step on any toes. BrightRoundCircle (talk) 11:57, 31 March 2017 (UTC)
  • I have been thinking on this debate further... and I have realized that while I very much disagree with S Marshall's rational for preferring the older "guarantee" language vs. the newer "reason" language... I actually agree with him when it comes to which version I prefer.Let me explain why:
I had actually forgotten that I was the editor who originally added this section to the policy, so I went back to examine why I felt it important to add in the first place (what my intent was). My original intent was indeed (as DonIago puts it) to say: "don't be surprised if your information gets deleted if all you can say about it is that it's verifiable". It was intended to be a more "after-the-fact" statement.
When I added this section, I was not thinking about BURDEN debates at all... the intent was to remind editors that there are lots of other reasons why verifiable information might be removed. Some are spelled out in policy and guidelines (for example, the addition might be too POV, etc)... but there are non-policy reasons as well (for example, other editors may simply not consider important enough to mention). Now... quite often, when someone has their addition removed (either for a policy reason, or for a non-policy reason), their first reaction will be to point to WP:V and say: "but my addition was verifiable... and WP:V says the requirement for inclusion is verifiability". So... we needed something in the policy that said "yes, but simply being verifiable isn't enough... it also has to pass a host of other policies and guidelines. AND... even if it passes every policy and guideline we have, it can still be omitted (if there is a consensus to do so)." Blueboar (talk) 13:49, 31 March 2017 (UTC)
I think you hold a unique position here in being able to phrase this in a way that would create a consensus and let the matter rest. BrightRoundCircle (talk) 14:51, 31 March 2017 (UTC)
I don't know exactly where this is going, but I'm neutral in the content / material word choice. North8000 (talk) 15:20, 31 March 2017 (UTC)
What needs to be verifiable are the claims and assertions made in the article. "Content" includes images and other media files, categories, templates, navboxes and other things that some editors like to edit-war about. In fact, we might want to empower editors to remove certain categories if they were unsourced (Category:Holocaust deniers anyone?) but I don't really want people putting citation needed templates next to maps or photographs of the subject.—S Marshall T/C 17:37, 31 March 2017 (UTC)
I find it a might difficult to imagine why a Category would need to be "sourced", or even how. Categories are not "content", they are structural conveniences. Any problems with them are addressed on the basis of other policies, guidelines, standards, or considerations; WP:V is not involved. It seems to me you are sliding off-topic here. ~ J. Johnson (JJ) (talk) 21:55, 31 March 2017 (UTC)
I think Blueboar's view is that of all-of-us-but-one, that verifiability is necessary but not sufficient for inclusion. That we haven't sorted out how to best explain that is because Marshall has been sitting cross-wise to the process. As he has been resistant to resolving this, perhaps it is time to allow that consensus isn't necessarily unanimity, and move on. ~ J. Johnson (JJ) (talk) 21:46, 31 March 2017 (UTC)
Not quite... My view is more along the line of "verifiability alone is not sufficient to prevent exclusion." Blueboar (talk) 22:09, 31 March 2017 (UTC)
And there we agree exactly.—S Marshall T/C 12:00, 1 April 2017 (UTC)
That is a rather awkward statement, with too many inversions (negations). While I am pretty good at logic, trying to sort out this sentence to see how it corresponds to our previous phrasings makes me wonder if someone slipped me decaf this morning. Note that the effect of the italicized words (prevent, permit, guarantee, etc., which are not necessarily exactly synonyms or antonyms) on our subject ("verifiability alone") can be inverted at two different places (is/is not, and ex-/in-clusion). Of the dozen permutations of this predicate with these three words, I challenge you all to determine which of the following are functionally equivalent, and which are functionally opposite.
1- verifiability alone is not sufficient to prevent exclusion.
2- verifiability alone is not sufficient to permit inclusion.
3- verifiability alone is not sufficient to guarantee inclusion.
4- verifiability alone is sufficient to guarantee inclusion.
5- verifiability alone is sufficient to permit inclusion.
We have quite enough struggle with variant interpretations of policy without makng them logic puzzles. We should use the simplest, most direct statement possible, avoiding all logical negations and inversions. ~ J. Johnson (JJ) (talk) 21:19, 1 April 2017 (UTC)

I'm starting to think it's a shame we can't address this as: "Verifiability is one element essential for inclusion of information in Wikipedia. Others include, but may not be limited to...." Or could we? DonIago (talk) 03:12, 1 April 2017 (UTC)

IMHO that would take a whole book on how Wikipedia works, and wp:ver would not be the place for it.  :-) North8000 (talk) 03:15, 1 April 2017 (UTC)

Another attempt at resolution

OK... I am beginning to think that we arguing just for the sake of argument. If I am reading the comments above correctly, most of us seem to feel that there is not a lot of difference between versions. While we may prefer one version over the other, we could live with either (I know this is my stance). Now, S Marshall has said that he would be satisfied if we simply returned to the older "Guarantee" language. So... does anyone actually object to doing this? If so, please state your concerns. Blueboar (talk) 11:57, 1 April 2017 (UTC)

Yes, i do object, and that is the main core of this entire discussion. North8000 (talk) 13:32, 1 April 2017 (UTC)
Since Blueboar and S Marshall both agreed above on the changing of "verifiability alone is not a reason for inclusion" into "verifiability alone is not sufficient to prevent exclusion" we can do that (assuming there's consensus), instead of throwing away the version everybody else in the previous discussion agreed upon. BrightRoundCircle (talk) 13:39, 1 April 2017 (UTC)
We should stat by clarifying the proposed change; I was confused from the above. The current wording is "However, while verifiability is required for including something, it is not a reason for inclusion." North8000 (talk) 13:45, 1 April 2017 (UTC)
Well, if I can parse the, erm, syntax in a few steps: "verifiability alone is not sufficient to prevent exclusion", means, "verifiability alone is not sufficient for inclusion", which means, "verifiability alone is not reason for inclusion", so that would just rid of the "a". Alanscottwalker (talk) 13:54, 1 April 2017 (UTC)
I know. BrightRoundCircle (talk) 14:01, 1 April 2017 (UTC)
The thing is... taking about verifiability being the "reason for inclusion" is not really relevant. No one actually adds information because it is verifiable. They add verifiable information because they think it is important, or useful, or (at least) interesting. The problem is that, sometimes, other editors don't think the verifiable information is important, useful or interesting enough to be kept. The information might be removed, even though it is verifiable. It is at this point (post-removal) that editors sometimes respond with: "but it is verifiable". This response only crops up after someone else removes text. So... I feel we need to phrase our caution (that information might be removed even when it is verifiable) in terms of responding to removals, not in terms of reasons for addition.
In any case, we have gotten off topic (again) ... I started this new thread because I wanted to find out who actually objects to the old "guarantee" language (and why). To that end... North has stated that he does object... so... NORTH, would you mind explaining (or perhaps re-explaining) why you object? Blueboar (talk) 14:47, 1 April 2017 (UTC)
Huh? When someone adds, for example, 'Fred tweets he thinks Mr. Politician is a so-and-so', they do add it because it is verifiable (unless, they are making it up) otherwise they would not know about it (also, in adding something like that, in a context, they may breach multiple policies and guidelines) and certainly not just that 'it's not useful'. Alanscottwalker (talk) 15:56, 1 April 2017 (UTC)
Blueboar, answering your question (somewhat repeating with but with an update):
  1. First, WP:ver in essence imposes a requirement for inclusion, nothing more. Nowhere does it say anything about meeting wp:ver being a full or partial reason for inclusion. The lede of the policy should be a summary of the policy. "Not sufficient" clearly implies that, per wp:ver, meeting wp:ver is, to a partial extent, a "reason" for inclusion. So, including "not sufficient" in the lead is in conflict with the body of the policy.
  2. Second, as detailed at length in the discussion on this, a common problem (rooted in a false false urban legend in Wikipedia) is mis-use or mis-reading of this policy to claim that meeting wp:ver is a force or reason for inclusion. For example, every time an editor says the ubiquitous "undo removal of sourced material" in an edit summary they are in essence making this claim. Nobody makes the claim that it is an absolute mandate for inclusion, but it is quite common to make the false claim that it is a strong force or reason for inclusion. This false urban legend is extensively used by wiki-lawyers to POV articles. Flatly saying that meeting wp:ver is not a reason for inclusion in the lede is (unlike "not sufficient") an accurate summary of of the body of the policy, and directly addresses the false urban legend.
  3. Third, IMHO many folks (including, going by memory, one or more current or ex Arbcom members) have said that Wikipedia has an overly byzantine set of confusing and overlapping / conflicting policies. This is a perfect example of sitting on the edge of that precipice, and avoiding the abyss. Regarding content, WP:VER sets a requirement for inclusion, nothing more unless we let it be a mess. It should not try to wade into the immensely complicated question of "sufficiency" for inclusion. To start with "sufficient" requires not being in violation of dozens of policies and core guidelines. If those are complied with, there is an immensely varying additional bar for inclusion. At the low end might be an uncontested inclusion at an obscure article where the additional bar is essentially zero. At the other end of a spectrum, the additional bar for inclusion (on a contested major item) might be getting a strong consensus in an RFC. WP:VER should stick to imposing the verifiability requirement, not start reflecting on or overlapping with the hugely complex area of what constitutes sufficiency for inclusion.
  4. Fourth, for new timid editors, this falsely implies that there is some unknown-to-them wiki-rule about needing a compelling reason to include something. This imposes a false hurdle, conflicts with new editor objectives (and wp:bold)
Sincerely, North8000 (talk) 20:42, 1 April 2017 (UTC)
I am confused... what I get from your explanation is that you strongly object to the word "sufficient" ... but that word isn't in the old "guarantee" version. Am I not understanding? Blueboar (talk) 20:58, 1 April 2017 (UTC)
Now I'm getting confused. I guess that the best answer I can give is that it's shorthand for S Marshall's proposal. I've been accepting the premise that S Marshall was arguing for a historic version, but now I can't see where anything like that existed. North8000 (talk) 22:09, 1 April 2017 (UTC)

Sigh, it seems I was not clear about what I was asking... so, let me start again, and I will try to be clearer. From February, 2013 to February 2017 the sub-section in question read:

Verifiability does not guarantee inclusion

While information must be verifiable in order to be included in an article, this does not mean that all verifiable information must be included in an article. Consensus may determine that certain information does not improve an article, and that it should be omitted or presented instead in a different article. The onus to achieve consensus for inclusion is on those seeking to include disputed content.

My question is: does anyone actually object to this language? If you do object to this language, please state what that objection is (as concisely as possible). Blueboar (talk) 00:53, 2 April 2017 (UTC)

I assume that you mean object to going back to that. I would object to going back to that because an improvement would have been lost, per my 4 reasons given above, but even more so because it might cause us to lose the succinct summary in the lede. But I'm getting worn out, I'm probably open to/ vulnerable to a bad compromise.  :-) North8000 (talk) 13:10, 2 April 2017 (UTC)
Thank you... yes, that answers my question. Anyone else? Blueboar (talk) 17:12, 2 April 2017 (UTC)
Yes. For starters, could we have strike-out and underline to highlight the differences? Otherwise we are likely to get (ha! have gotten) into side-issues. The issue here is not a general "how is this language?", it's a matter of two specific formulations. And let's not forget we are talking about two different but related places (four formulations?), so we need to consider both.
I also object to any further consideration of this matter at this time. Like North8000, I'm getting worn out. I think we all are fatigued with this discussion, and vulnerable to bad thinking. (On that basis alone I'm inclined to oppose any changes at this time.) While I appreciate your attempt to resolve this, this discussion has gone on so long, and been so torturous, that no one can properly summarize the main points, or even key points. I suggest we close this discussion as being hopelessly entangled, and take a break. While the rest of us cool our brains you might work with Marshall to determine what his conerns are, and hopefully with some examples. You might also consider some of the questions and objections the rest of us have raised, and work with Marshall on how to best respond. Then you (not Marshall!) could raise a question as to aptness of the current language.
If we must continue at this time, then I also object to (as you suggest at the top of this section) any return to the older "guarantee" language. The objections to guarantee were previously aired at #Where verifiability is not an adequate criterion (13:44, 8 Feb; 23:31, 8 Feb; 22:56, 10 Feb; 22:19, 12 Feb). Do we need to rehash that? ~ J. Johnson (JJ) (talk) 22:29, 2 April 2017 (UTC)
A break sounds like a fine idea... and in the meantime, I would be happy to work with anyone behind the scenes (as it were). Yes, I have a preference for the older language, but that preference is not so strong that I object to the newer... so my goal is simply to find language we all can agree on (if possible). Blueboar (talk) 22:46, 2 April 2017 (UTC)
A break sounds like a good idea. I know from the past that S Marshall (like the other good folks who participated in this) is driven by wanting to end up with a quality encyclopedia, even though I disagree with him on this topic. North8000 (talk) 23:06, 2 April 2017 (UTC)
The differences are minor but the new version is still an improvement. A break would be good, and maybe later this could be tackled issue-by-issue instead in an all-or-nothing revert. Bright☀ 10:47, 3 April 2017 (UTC)
We're stuck and a break will achieve nothing except further entrenchment; it's quite clear that nobody's position will change in a few weeks' time. I've delayed launching the RfC to allow time for a killer suggestion, but there is none. I see no good reason why I shouldn't proceed to put the question to the community, and I'm uninterested in J Johnson's opinion about who should be the one to ask.—S Marshall T/C 21:55, 3 April 2017 (UTC)
Since Marshall can't (or won't) see what a total hash he has made of "putting the question" so far, won't engage in consensus building, and rejects my suggestion that he would be more successful letting someone else advocate on his behalf, it is quite doubtful that proceeding with an RfC would be more productive or less disruptive than what we have seen to date in this discussion. So if someone else could persuade him to not run off the cliff of an RfC, this would be a most good time to do so. ~ J. Johnson (JJ) (talk) 00:29, 4 April 2017 (UTC)
Whatever happens, it'll be over in 30 days, so let it happen. The discussion has already spanned two weeks and got nowhere, might as well give it another month just so it can be done. Bright☀ 10:34, 4 April 2017 (UTC)
I like the break idea. North8000 (talk) 12:24, 4 April 2017 (UTC)

Before we either take a break, or go to an RFC... I want to suggest a compromise. We have been debating this in terms of it being an either/or choice between two versions... but does it need to be so? Instead of choosing between "Verifiability is not a reason for inclusion" and "Verifiability does not guarantee inclusion", perhaps we should say BOTH. Something along the lines of: "Verifiability is not a reason for inclusion, nor is it a guarantee of inclusion". If everyone is too worn out to discuss this idea now, I do understand... but please think about it. Blueboar (talk) 12:33, 4 April 2017 (UTC)

  • The purpose of our encyclopaedia is to present verifiable information to the reading public. Verifiable information should be added to the encyclopaedia, and this policy should encourage and enable that. I think that verifiability is a reason for inclusion; and I simply disagree with those who say otherwise. This novel, dramatic and very unwise change to the policy should not have happened without a 30-day community RfC. If you don't trust me to start that RfC, start it yourselves.—S Marshall T/C 16:31, 4 April 2017 (UTC)
My neighbor's house is in the Gothic Revival style of architecture. This is verifiable (multiple real estate listings say so). However, my neighbor's house is not a notable example of the Gothic Revival style. Must we mention my neighbor's house and the verifiable fact that it is in the Gothic Revival style in Wikipeida? If not, then verifiability is not a "reason" for inclusion (because we just excluded it). Blueboar (talk) 19:28, 4 April 2017 (UTC)
Verifiability isn't a guarantee of inclusion. The fact that something's verifiable suggests we should include it, unless it fails another core policy (in the case of that hypothetical addition, WP:NOT would kill it off).—S Marshall T/C 19:38, 4 April 2017 (UTC)
We might wish that people would have all the policies in mind simultaneously as they add something to an article, or remove it, but in many cases, editors will think about (1) writing a good article and (2) one or two of the policies that seem particularly relevant. So although this policy should mesh nicely with the other policies, it would be better if the wording leaned toward self-sufficiency rather than became nonsensical if one wasn't simultaneously thinking of all the other policies while reading this one.
Trivial, unencyclopedic information fails WP:NOT, but this policy should recognize that including that kind of junk is just bad writing. Thus, this policy shouldn't be say everything verifiable that doesn't fail a policy should be included, because that elevates the WP:NOT policy above the principle that editors should try to write well. Good writing is the reason we don't include uninteresting trivia, and WP:NOT just gives good writers the authority to fix bad writing. Jc3s5h (talk) 20:08, 4 April 2017 (UTC)
I would go further, and say that this policy (WP:V) should not be saying anything (nor be invocable) about inclusion or exclusion on any other basis, or any other policy, beyond "verifiability in reliable sources is required". Period. That is the full scope of WP:V, and let the other policies speak for themselves. What concerns several of us here is (e.g.) questionable material not covered by WP:NOT and other policies, where there may be a majority sentiment it is inappropriate, but some contrary editor insists there is no rule against it, and we can't remove it without a full-fledged RfC to determine community consensus. Which is, effectively, the argument that verifiability alone guarantees inclusion (up until Rfc'ed otherwise).
As to 30 more days of this: gah. An RfC immediately following this discussion will inevitably incorporate this failed discussion, with all of its confusions and ramblings. We will be running with vague and confused recollections of what was said, tempers will fray, and we will just dig deeper holes. We need to back off, cool off, clear away the rhetorical rubble, and start fresh. As well as start from ground level. E.g.: what are the concerns? what are the examples? what are the pertinent considerations? etc. But not now. Let's have a break. About three weeks seems good. ~ J. Johnson (JJ) (talk) 00:09, 5 April 2017 (UTC)
Agree.North8000 (talk) 12:45, 5 April 2017 (UTC)
It appears we all (or all but one?) agree to take a break on this question of phrasing. I suggest someone close this thread (so anyone passing by doesn't accidentally re-ignite it), and come May we can start a new discussion. And someone please work with S Marshall to distill his concerns into some kind of definite form that can be usefully addressed. ~ J. Johnson (JJ) (talk) 20:19, 7 April 2017 (UTC)
  • I've explained to you before that you don't control whether I can start an RfC. I see no reason for delay except to allow this inferior language to become more entrenched, so I'll be starting the RfC later today.—S Marshall T/C 14:02, 9 April 2017 (UTC)
A strawman argument, as I have never claimed to control you in any way. Although your petulancr when I make a good-faith suggestion for your own advantage suggests a possible point of leverage. ~ J. Johnson (JJ) (talk) 22:07, 9 April 2017 (UTC)

Just in mainspace?

The current version doesn't seem to indicate clearly if it applies to pages outside of the main space or not. Could a note be added to clarify this? Siuenti (씨유엔티) 08:22, 7 April 2017 (UTC)

In the lead it says "all material in Wikipedia mainspace, including everything in articles, lists and captions, must be verifiable." It is also mentioned in the nutshell that it is articles that must be verifiable. Jc3s5h (talk) 11:29, 7 April 2017 (UTC)
One might take " its content is determined by published information rather than the beliefs or experiences of its editors" to imply that it does apply more widely, so I think it would be best to add a note on the lines of " (This policy of no original research does not apply to talk pages and other pages which evaluate article content and sources, such as deletion discussions or policy noticeboards.)" from over the way. Siuenti (씨유엔티) 11:52, 7 April 2017 (UTC)
Do we really need to say it? Are we actually getting disputes where people are arguing that WP:V applies outside of mainspace? I am not opposed to adding something like what Siuenti suggests, but it does strike me as unnecessary instruction creep. Blueboar (talk) 12:52, 7 April 2017 (UTC)
I was recently told off in rather florid terms for my "original research" leading to the determination that a source failed WP:MEDRS, on the discussion page of an article, even though WP:NOR very explicitly places discussion pages outside its jurisdiction. Indeed, no straw seems safe from the clutches of single-purpose editors with entrenched interests, even those that we believe have been safely stowed away behind black-letter policies. Sławomir Biały (talk) 13:03, 7 April 2017 (UTC)
@Sławomir Biały: would people be able to double-check that research? Siuenti (씨유엔티) 13:08, 7 April 2017 (UTC)
Certainly. The source was not MEDRS, and raised red flag claims. It was specifically the red-flagginess of the claims that was questioned as OR. Sławomir Biały (talk) 13:16, 7 April 2017 (UTC)
So one user says "red flaggy because X Y Z" and one says "that's OR you can't say that"? Siuenti (씨유엔티) 13:21, 7 April 2017 (UTC)
Right. Sławomir Biały (talk) 13:23, 7 April 2017 (UTC)

So it seems to me that people can easily interpret WP:NPOV and WP:V as applying where they don't, or at least not be sure if they apply or not. So clarification would be good. Siuenti (씨유엔티) 13:28, 7 April 2017 (UTC)

We should not discourage asking for and focusing on sources. If most can come to agreement on where sources lead, they are likely to have lasting consensus. Sourced arguments should be seen as helpful and encouraged, not discouraged. That's not to say everything one argues needs a source - nonetheless, to elevate a discussion between position x and y (where assertions and worse fly), a good way often is to see which position can be better sourced. Alanscottwalker (talk) 13:30, 7 April 2017 (UTC)
I think it is a mistake to regard the better sourced position by default as the better one. For example, suppose someone writes up a new theory of gravitation, which is not reported on by secondary sources. The fact that no one besides the originator has commented on the new theory should not be taken as evidence that it carries any implicit advantage in a discussion by virtue of being "better sourced". Part of what we need to do on discussion pages is assess some sources absent sources that take a contrary view. Sławomir Biały (talk) 13:41, 7 April 2017 (UTC)
So a disputed statement should at least have some kind of source? Siuenti (씨유엔티) 13:42, 7 April 2017 (UTC)
No. Absolutely not. See below. Sławomir Biały (talk) 15:20, 7 April 2017 (UTC)
If 'someone writes up a new theory of gravitation, which is not reported on by secondary sources' - that is focusing on sources, the focus there being no secondary source cares about it, if people can agree that no secondary source cares about because, we consulted and asked for secondary sources, you have reached a consensus. Alanscottwalker (talk) 13:53, 7 April 2017 (UTC)
So yeah, you had a dispute and you looked at sources to resolve that. Someone might think that was how you resolved all your disputes whatever the namespace. Siuenti (씨유엔티) 14:01, 7 April 2017 (UTC)
Yes, the issue here is not that our discussion focused on sources and found that there were no secondary sources. The issue is that we did not have a source that told us this. The statement "there are no secondary sources that verifies XYZ theory" therefore does not satisfy WP:V. Meta-discussions about sources do not and should not require sources, and I do not think it would harm the project to point this out. Sławomir Biały (talk) 15:11, 7 April 2017 (UTC)
Meta-discussions about sources are still based around sources or the lack thereof. Siuenti (씨유엔티) 15:32, 7 April 2017 (UTC)
I think you miss the point. Being "based around sources" is different from "sources say XYZ". When we write something in an article, it is because a source affirmatively says it (WP:V). We can write something on the discussion page that "there is no source for XYZ" (which fails WP:V). Do you not see the difference? Sławomir Biały (talk) 15:35, 7 April 2017 (UTC)
If I said "Wikipedia bases its decisions on sources", statements like "there is no source for XYZ" would not be incompatible with that. That's why it would be reasonable to think WP:V implies that, even though things could get difficult if it was applied recursively Siuenti (씨유엔티) 15:48, 7 April 2017 (UTC)
But do you feel that it is appropriate to state in a Wikipedia article that "there is no secondary source for Professor X's" theory? Or does WP:V imply that we need a source for that? Sławomir Biały (talk) 16:02, 7 April 2017 (UTC)
If there was one would you definitely have found it? Siuenti (씨유엔티) 16:10, 7 April 2017 (UTC)
If you aren't sure maybe "no secondary sources are known" Siuenti (씨유엔티) 16:15, 7 April 2017 (UTC)
I am asking you, as a rhetorical device, since you apparently do not understand the examples that I gave. Do you believe that "no secondary sources are known" is an appropriate WP:V-compliant statement for an article to make, without citing a source? Thanks, Sławomir Biały (talk) 17:01, 7 April 2017 (UTC)
If you made a good-faith effort to look for them, yes. Siuenti (씨유엔티) 17:27, 7 April 2017 (UTC)
Good to know, but I expect this is not an interpretation of policy that enjoys wide support. Sławomir Biały (talk) 17:30, 7 April 2017 (UTC)

Do I understand correctly that someone is claiming that, in a discussion, one cannot give one's own opinion as to whether or not a source is a reliable source? That the determination of whether ABC Monthly is a reliable source must be based exclusively on reliable meta-sources that say things like "ABC Monthly is a reliable source under Wikipedia's definition of that term" or "ABC Monthly completely fails to meet Wikipedia's criteria for reliable sources"? As though such meta-sources exist. Besides, how would we evaluate the reliability of the meta-sources? Turtles all the way down. Largoplazo (talk) 16:22, 7 April 2017 (UTC)

As I understand it, the reason WP:V doesn't say anything about its applicability outside the main space, is that people would start claiming that. Siuenti (씨유엔티) 16:26, 7 April 2017 (UTC)
It's a commonly used tactic....to imply that a talk page comment is illegitimate unless it is sourced, and to stifle talk page comments by saying that someone needs to source comments in order to make them. Perhaps a more direct mention would be good. North8000 (talk) 16:34, 7 April 2017 (UTC)
People sometimes logjam up discussions with statements that aren't true, and don't provide evidence that they are true. I think if two editors reasonably dispute a fact evidence should be forthcoming or the assertion put on the shelf. Siuenti (씨유엔티) 16:50, 7 April 2017 (UTC)
Well, that has nothing to do with WP:V. It isn't question of Wikipedia policy. It's the same principle that applies in any discussion/argument/debate in the world. If you claim something is true and someone else doesn't believe you, then the burden of proof may fall on you (if you're the one making a specific claim), not because there's a policy that you have to conform to, but to establish the credibility of your claim to the satisfaction of whoever you're debating with. Otherwise, they are under no obligation to agree with you, if they have reason to believe otherwise. The alternative is to decide you don't care whether the others agree with you, and drop out of the discussion.Largoplazo (talk) 16:59, 7 April 2017 (UTC)
If I say "let's do this because A is true", and Bob says "no, A is not true", this issue should be decided by evidence, not by Bob's say-so. Siuenti (씨유엔티) 17:08, 7 April 2017 (UTC)
That's what I just said, and that has nothing to do with conforming to Wikipedia's WP:V policy. Obviously, since the concept of burden of proof in an argument existed for millennia before there was such a thing as Wikipedia. Largoplazo (talk) 17:13, 7 April 2017 (UTC)
If I say the other guy has the burden of proof does that mean I can say his objection doesn't count in terms of determining consensus? Siuenti (씨유엔티) 17:25, 7 April 2017 (UTC)
If you aren't convinced, then you aren't convinced. You aren't Wikipedia, and your standards for believing something someone tells you, or becoming convinced of it, may be different from what WP:V provides. But you, like everyone else, should be evaluating input rationally and in good faith. In the case of Russell's teapot, it's reasonable to ask someone who claims that a teapot is in orbit in some part of the solar system to prove it. It's absurd for the person making that claim to ask those who don't believe him to prove that it isn't true. Largoplazo (talk) 20:44, 7 April 2017 (UTC)
Hmm. If I think there is no such thing as Russels teapot, it would perhaps be reasonable for me to ask for an RS and assume it doesn't exist until that RS appears? Siuenti (씨유엔티) 20:55, 7 April 2017 (UTC)

Re wanting to state in an article that there are no sources to support such-and-such a claim: Uh ... is this still about the kid with Asperger's syndrome who supposedly has solved all the deepest mysteries of the universe? Largoplazo (talk) 16:43, 7 April 2017 (UTC)

Did anyone here articulate a position "wanting to state in an article that there are no sources to support such-and-such a claim"? That appears to be the position of User:Siuenti, who has no apparent connection with the article you're most likely thinking of. But I might be wrong about that, in which case this just seems like a red herring. Sławomir Biały (talk) 17:28, 7 April 2017 (UTC)
Not sure whose on which side. If someone say that I wouldn't take it out. If someone else took it out I wouldn't put it back. Siuenti (씨유엔티) 17:31, 7 April 2017 (UTC)
Then it's just a bizarre red herring. To clarify my own position, it is that WP:V does not apply to talk pages (obviously), and less obviously, it is not always the "sourced position" that should be given more weight. "Professor X published a new theory in the academic journal Journal of Professor X." The talk reply "That's not a real journal" most emphatically doesn't require a source. Neither does "That's a fringe theory that no one has commented upon." Yet the latter are strong arguments for exclusion, despite having no sources to back them up. Sławomir Biały (talk) 17:45, 7 April 2017 (UTC)

The thing is, BURDEN is a part of this policy and it does demand that one fulfill their BURDEN, it makes little sense to not discuss the sources that fulfill that burden on talk pages. And Talk Page guideline focusing discussion on sourced assertions, not editors' assertions, is reinforced at WP:TALK#OBJECTIVE, WP:TALK#FACTS and WP:TPYES. (Although, I do agree that discussing whether something is reliable enough for something is a judgement call, it still is informed by focusing on the (verifiable) facts of the source (eg, what are the circumstances of its publication) - which is focusing discussion on sources). -- Alanscottwalker (talk) 17:52, 7 April 2017 (UTC)

I agree, focusing on sources is something that should be encouraged. But it is a mistake to infer that WP:V itself should be followed in metadiscussions of sources. Any view that editor judgment is automatically trumped by "sources" in a discussion is deeply wrong. Sławomir Biały (talk) 18:05, 7 April 2017 (UTC)
WP:NOTPART -". It is therefore not necessary to provide reliable sources to verify Wikipedia's administrative pages, or to phrase Wikipedia procedures or principles in a neutral manner, or to cite an outside authority in determining Wikipedia's editorial practices".--Moxy (talk) 16:59, 8 April 2017 (UTC)

What kind of statements are in for WP:V?

I have a strong suspicion that this long discussion is because different people are thinking/talking about different things which may be uttered in talk page. Therefore let me start from some self-made taxonomy.

  • SoF Statement of fact
    • requires citation on demand regardless space. A statement of an opinion (SoO) of a person other than a wikipedian is a fact as well (to be distinguished from the fact(s) stated in the opinion (apply recursively :-): "A said that B said that X is an idiot").
  • SoN Statement of negation ("There are no blue monkeys")
    • in article space admissible only if from WP:RS, because it is always an issue of falsifiability.
    • Admissible in talk pages provided due diligence
  • SoWO Statement of wikipedian's opinion
    • Disregarded in article space WP:NOR
    • May be disputed/disagreed in talk space, but may serve as an insight "in search of truth".
  • SoWOboF Statement of Wikipedian's opinion allegedly based on facts the Wikipedian inferred from sources
    • A variation of the Statement of argument below.
  • SoA Statement of argument ("since A therefore B")
    • WP:NOR admits a limited amount rather trivial and self-evident logic and math, and even such may be subject to opposition
    • talk pages: Fully admissible during a discussion. May be objected by a counter-argument. And the ultimate argument must be either logical, or verifiable, or based on our guidelines, including WP:COMMON.
  • SoB ("You are an idiot/racist son of a bitch/etc")
  • Any other takes?

Now IMO it is better seen the scope of WP:V in talk pages. Preliminary conclusion: the issue is perfectly handled by WP:COMMON and hence talk pages should not be discussed in WP:V. Staszek Lem (talk) 21:13, 7 April 2017 (UTC)

Only question I have about your taxonomy, is in SoWObof/SoA: If I argue in the form 'facts inferred, therefore conclusion', should I generally be willing to provide source(s) for those facts (perhaps that is covered by your SoF)? As an aside, Common, at least the one you linked to, is not a guideline, but there is a brief reference to appeal to common sense in CONSENSUS policy. -- Alanscottwalker (talk) 21:37, 7 April 2017 (UTC)
facts inferred -yes as in any SoA you must check both premises and logic. As for "Common", talk pages are basically not governed by strict policies besides NOTAFORUM, WP:AGF, and CIVILITY. So IMO "use common sense" is a reasonable advice. Staszek Lem (talk)
Something that I think is left out is that although the policy is WP:V, "verifiability not truth", it is also the case that WP:TRUTHMATTERS, and ultimately the verifiability policy is a proxy for "the truth" (at least, one hopes). Something may be demonstrably true without being either WP:Verifiably true or logically true (in the sense of propositional calculus). For example, suppose that the statement "Professor X has not published a paper in quantum chromodynamics" is a true statement about the world, but for which there are no secondary sources commenting on Professor X's lack of publications in that area. While we cannot cite sources for that, it is nontheless easily established using Google Scholar or other database that indexes all published papers in the field of quantum chromodynamics. So the truth of that statement is not something that comes from a source, and nor is it something that comes from a logical argument, but rather is inferred from a knowledge of the workings of the scholarly world. Such a statement would be a perfectly acceptable argument on a talk page, yet it does not apparently fit into your taxonomy. Sławomir Biały (talk) 22:10, 7 April 2017 (UTC)
What I am not getting about some of your points, is that what you talk about can well be stated as a fact and sourced: 'In these indexes, professor has no record of publishing. (Source these indexes).' Alanscottwalker 22:21, 7 April 2017 (UTC) (Fixed sig. Alanscottwalker (talk) 22:42, 7 April 2017 (UTC))
Re: easily established using Google Scholar - for the purpose of talk, yes (e.g to establish credentials for the purposes of WP:RS), for the purpose of article text, hell no way; discussed countless times. Staszek Lem (talk) 22:36, 7 April 2017 (UTC)
Sure. But it's not because the statement that 'Google scholar contains none' is unverifiable (all that is needed to verify that is the proper search), it's because for article space, it's too trivial (NOT), or too irrelevant (UNDUE), or too weakly sourced (OR/primary), or some combination thereof and more. -- Alanscottwalker (talk) 22:59, 7 April 2017 (UTC)
Sorry, you missed the point (probably because of my insufficient quoting). In question was the truthfulness of "Professor X has not published a paper in quantum chromodynamics". And for this purpose, the verifiability of the statement about Scholar has the value as I described. Staszek Lem (talk) 23:16, 7 April 2017 (UTC)
Oh, sorry, my fault, but that highlights another point about the good of discussing and producing sources on talk pages. It's hard to begin to determine what is supportable without producing the source. Alanscottwalker (talk) 23:31, 7 April 2017 (UTC)
So "verifiable" doesn't mean WP:Verifiable. Sławomir Biały (talk) 11:01, 9 April 2017 (UTC)
100% of this problem would be fixed if people maintained a clear distinction between personal opinion and fact. In my opinion, anyway... Guy (Help!) 22:13, 7 April 2017 (UTC)

What an excellent analysis. Y'all are opening up about 5 big important topics at once. Where's that going to lead?  :-) Sincerely, North8000 (talk) 01:29, 8 April 2017 (UTC)

"Statement of fact requires citation on demand" and "Statement of negation ... admissible in talk pages provided due diligence" look very very good to me, because they are the kind of thing that can completely derail a discussion when people feel no obligation to cite/perform due diligence for them due to their "expertise" on the matter. Siuenti (씨유엔티) 23:13, 8 April 2017 (UTC)

I partially agree with what Blueboar wrote above. This policy is for article space. Trying to fit it around other spaces would be instruction creep, but unlike Blueboar I am against "adding something like what Siuenti suggests". The process for editing article space other spaces vary, trying to fit explanations for both would bloat the policy page. This discussion is a good example of why WP:V is not appropriate, let alone the fun someone could have at WP:ANI with WP:V restrictions on those discussions.

There are plenty of discussions about subjects on the talk page that do not fit easily into the "self-made taxonomy" user:Staszek Lem. Two examples: WP:RM, and any discussion like this one that relies on producing facts from the Wikipdia space eg, WP:V is an unreliable source as far as WP:V defines it, so it can not be used to verify a fact under its own definition (WP:CIRCULAR). Now I am sure that we could have fun arguingdiscussing if I am wrong about my interpretation of WP:CIRCULAR, but if one really really wants to extend WP:V to cover other spaces then that is a section would need to be rewritten for clarity. Another example from WP:V is "Any material lacking a reliable source directly supporting it may be removed and should not be restored without an inline citation to a reliable source". One can not delete another editors comment on a talk page or at ANI just because one think it lacks "a reliable source directly supporting it". -- PBS (talk) 01:23, 9 April 2017 (UTC)

Um... Let me clarify my statement... because PBS seems to think I support having WP:V apply outside of mainspace. No. I think adding anything on this is unnecessary instruction creep, but I agree that WP:V does not apply outside main space. Sorry if I was unclear. Blueboar (talk) 03:08, 9 April 2017 (UTC)
So outside of mainspace, people can make statements like "the moon is made of cheese" or "there are no blue monkeys" and not bother to provide a citation or due any diligence when challenged. Siuenti (씨유엔티) 06:40, 9 April 2017 (UTC)
Sure. What's more, people can propose that WP:V apply to all pages despite no other editor agreeing with them. However, chatting on an article talk page about whether blue monkeys might exist would almost always be a violation of WP:NOTFORUM. There are occasional NOTFORUM exceptions for long and tedious discussions where someone makes a useful joke that leads to a small number of off-topic exchanges—that can be useful for community building. This page shows that discussing blue monkeys might be helpful for a while, but in general any discussion on an article talk page about content that has no reliable sources and which violates WP:REDFLAG is off-topic and should stop. Johnuniq (talk) 07:19, 9 April 2017 (UTC)
Suppose the question of what kind of dairy product the moon is made out of, is relevant to the content of the article? Siuenti (씨유엔티) 07:23, 9 April 2017 (UTC)
I think that is far too academic a question for this page! See the entry for Moon in The Oxford Companion to Cheese (2016).
  • Catherine Donnelly, ed. (2016). "Moon". The Oxford Companion to Cheese. Oxford Companions. Oxford University Press. p. 493. ISBN 9780199330881.
--PBS (talk) 08:10, 9 April 2017 (UTC)
Come on guys, competence please. "The moon is made of cheese" is just a placeholder for an objection to an improvement. Siuenti (씨유엔티) 08:19, 9 April 2017 (UTC)
It is often surprising what already exists in Wikipedia! There is already an article on "The Moon is made of green cheese" — as there is for Turtles all the way down  .--PBS (talk) 08:25, 9 April 2017 (UTC)
Please don't take this off-topic. Someone says "the article shouldn't change because X". They provide no evidence for X, and evidence has already been provided against X. Should their objection prevent changes to the article, which is under discretionary sanctions. Siuenti (씨유엔티) 08:33, 9 April 2017 (UTC)
As with the discussion on Talk:Liancourt_Rocks#RFC on changing the title? This seems to me to becoming a fishing expedition. If I had not been the person who requested the move to its current name 10 years ago, I would have closed that RfC as an admin under the discretionary sanction of that page, (1) because the correct process to discuss name changes is WP:RM, and (2) it should be closed now due to snow as the vast majority of opinion expressed is "Status quo". I suggest that you drop your quest both here and there. -- PBS (talk) 08:58, 9 April 2017 (UTC)
This seems to me to be stonewalling. I suggest you start answering question instead of bringing up unrelated issues. Siuenti (씨유엔티) 09:07, 9 April 2017 (UTC)
It might help if you gave us a more specific issue to discuss, instead of a hypothetical. Was there a dispute at an actual article that inspired the question? If so, what article? Blueboar (talk) 10:58, 9 April 2017 (UTC)
At Liancourt Rocks I was trying to suggest tests which are not subject to cherry picking would be a good way of deciding the issue. I was informed there is no such test. At Donald Trump I was informed that the statement "Barack Obama is the 44th President of the USA" is false. In cases such as these I would like to be able to request evidence or diligence to support these statements. Siuenti (씨유엔티) 11:08, 9 April 2017 (UTC)
As to the Liancourt rocks, "there is no such test", just from your description, does have an implied source: relevant policies/guidelines. As to the Barack Obama issue, that appears from just your description to be a matter of context, and attention to context demonstrates diligence. So, if all you want is source and diligence, you have been given it. You are free to counter with relevant source(s) and/or diligence. Alanscottwalker (talk) 11:44, 9 April 2017 (UTC)
At this time I think you are telling about blue monkeys in the form of "relevant policies/guidelines" and I challenge you to go find them and show them to me. I don't really see how the context of a statement like ""Barack Obama is the 44th President of the USA" determines whether or not it is true. Siuenti (씨유엔티) 11:55, 9 April 2017 (UTC)
In your position, I would not draw attention to Talk:Liancourt Rocks#RFC on changing the title or Talk:Donald Trump/Archive 56#Why is "incumbent" necessary in the hatnote? because what is seen there is very misguided. Johnuniq (talk) 11:56, 9 April 2017 (UTC)
In your position I would not go look at Donald Trump and see if "incumbent" is in the hatnote and "current" is in the lede. Siuenti (씨유엔티) 12:16, 9 April 2017 (UTC)

Status Quo

So the status quo regarded disputed statements of fact on talk pages appears to be:

  • Neither side can require the other side to provide evidence
  • Regardless of any evidence provided or not provided, each POV has equal weight

Does that about sum it up? Siuenti (씨유엔티) 17:33, 9 April 2017 (UTC)

  • Comment-Everyone in a complex system has a slightly different interpretation of different things.But this one is stunning.Interpretation of the month??Winged Blades Godric 06:50, 10 April 2017 (UTC)
You are putting words into peoples mouths, and in you summary are contradicting various polices. See for example
  • WP:RM "The debate is not a vote; please do not make recommendations on the course of action to be taken that are not sustained by arguments."
  • WP:AfD "The debate is not a vote; please do not make recommendations on the course of action to be taken that are not sustained by arguments. When making your case or responding to others, explain how the article meets/violates policy rather than merely stating that it meets/violates the policy."
Administrators closing such debates will disregard opinions that do not meet these requirements. -- PBS (talk) 13:02, 10 April 2017 (UTC)

New problem with Archive.org

FYI, I started a new discussion today about a worrisome new problem with Archive.org that could make verifiability much more difficult for us.Anythingyouwant (talk) 21:29, 9 April 2017 (UTC)

Never mind, it's all resolved.Anythingyouwant (talk) 15:45, 10 April 2017 (UTC)

Citing inappropriate things

If I believe something which has no citation wouldn't be appropriate even if it did have a citation, why should I waste my time looking for a citation so it can be preserved ? Don't I have anything better to do? Siuenti (씨유엔티) 14:00, 11 April 2017 (UTC)

From WP:PRESERVE: "Preserve appropriate content. As long as any facts or ideas would belong in an encyclopedia, they should be retained in Wikipedia." (emphasis added) And for content that is verifiable, you would look to other policies and guidelines than WP:V to judge when it is not appropriate. postdlf (talk) 14:36, 11 April 2017 (UTC)
The sentence in the WP:BURDEN section that you refer to is qualified by its context, in particular the previous one: " it is advisable to communicate clearly that you have a considered reason to believe that the material in question cannot be verified." The request to "provide an inline citation yourself" applies only to the case where you want to remove content for being unverifiable, not when you think it is inadequate for other reasons. Diego (talk) 14:40, 11 April 2017 (UTC)
So I propose this version would be clearer:
If you think the material, if verifiable, is also appropriate, please to try to provide a citation yourself before considering whether to tag or remove it.
Siuenti (씨유엔티) 12:25, 12 April 2017 (UTC)
That version changes the meaning of the policy though, which is "even if you think that the material is not verifiable" (in the previous sentence), please try to provide a citation before removing it. The idea is that the editor may be surprised to find out that it was easy to verify after all. This minimum reality check is intended to prevent "trigger-happy" removal of valid content, that could create an inordinate workload on writers if it was not balanced with at least a cursory check prior to removal. Diego (talk) 13:33, 12 April 2017 (UTC)
re: why should I waste my time - you should not. Rather, you should look at the PRESERVE policy in full, which gives you more options: "consider cleaning up the writing, formatting or sourcing on the spot, or tagging it as necessary". In fact a bit below it gives you a long list of bullets "[i]nstead of removing content from an article". Of course, wikipedia is a volunteer project, nobody can tell you you should do something. Instead, the policy says what you should not: thou shalt not delete a text because you think something is wrong with it; admit that any of us (including you) may be mistaken. Fix it if you can; tag it if you can not, to warn both idle readers and busy wikignomes; remove it only as a last resource, after due diligence executed to verify it is indeed false or otherwisegrossly inappropriate. Staszek Lem (talk) 18:18, 12 April 2017 (UTC)
one small flaw with stating PRESERVE as a "Thou shalt not", Staszek... you (and the EP policy) include "clean up the writing" as an acceptable option... but quite often "cleaning the writing up" involves deleting text. In other words, it is OK to not preserve... as long as you replace with something else. Blueboar (talk) 21:00, 12 April 2017 (UTC)
"Deleting text" is vague: this may mean either deleting some words (without changing the overall meaning) or deleting information. Some people may be quite aggressive with rewriting, and it is not OK not to preserve information, especially while putting "copyedit" in edit summary. Staszek Lem (talk) 02:26, 13 April 2017 (UTC)
I agree that it is misleading to put an edit summary of "copy edit" when you completely rewrite a paragraph or section (or even an entire article) - but we are still allowed to completely rewrite. And sometimes information gets omitted in the process. It's called editorial discretion. Blueboar (talk) 11:10, 13 April 2017 (UTC)
"Thou shalt not" is not consistent with "consider instead" - also, such commands as "thou shalt not" are incredibly rare in our policies for many reasons. Note, when something is poorly sourced or unsourced, it is often an issue of not just V but of relevance, and unclarity or other policies - Although I like PRESERVE, the way it's sometimes talked about seems to skip over some things rather glibly, like, it regularly can be hard to even begin to judge "appropriateness" without the source. In addition, it's almost a request to get into the author's mind (what were they thinking), and we should admit that the mode of preserve -- its process -- is rather backwards from what NPOV, and OR (and even parts of V) would suggest the way to do things are - the article sentence should follow what the sources say, not the sources follow the sentence. Alanscottwalker (talk) 13:21, 13 April 2017 (UTC)

How does one transparently cite more than one part of an e-book in a single citation?

(Background.) When we want to verify information that we've derived from more than one page of a printed book we simply provide the page numbers. That might be a page range, but it might also be two or more pages that are not nearby each other.

Non-pdf e-books don't have pages. Most devices used by readers will break the book up into the equivalent of a page view, but when viewed as between different devices, an entirely different number of "pages" may be shown.

Still, so long as the information comes from one location in an e-book, it seems to me providing the URL that takes the person to that exact part of the source provides transparent verification, right? But what about when we have the equivalent of the need to cite two or more (discontinuous) "pages" from an e-book for a single citation; how do we provide a transparent citation that targets the two locations in it?

The only way I can see to do this is to provide two separate URLs in the cite (which AFAIK, cannot be done using our citation templates—can it?), or, really "clunkily", cite the same source twice in a row, with identical details but for the URLs targeting the two separate locations in it. Has this been discussed and is there any solutions or recommendations? Our need to cite e-books is only going to increase as time goes by.--Fuhghettaboutit (talk) 22:22, 26 April 2017 (UTC)

  • I think best practice would be to include both links in the same reference, but how to go about doing that within our citation templates is an interesting question. @Headbomb: may be interested and is usually helpful with this kind of thing.—S Marshall T/C 23:30, 26 April 2017 (UTC)
Use {{rp}}.

The sun is a star.[1]: sun The moon is not a star.[1]: moon

Although I would not use URLs at all, and simply give the name of the chapter in the {{cite}} template to avoid all of this.

The sun is a star.[1]: chapter 1  The moon is not a star.[1]: chapter 8 

References

Bright☀ 08:50, 27 April 2017 (UTC)
Aaaarrrgghhf! No, no, not {{rp}}; it's ugly! it's too horrendous to be tolerated!!
Folks, look, the seeming difficulty here is more conceptual than real. And is readily handled. But there needs to be clarification of some concepts.
First: URLs are not citations. You can't just point somewhere and say: that's it. Proper citations (here, as well as well-established practice elsewhere, print and non-print) should give enough bibliographic detail to describe and identify the source – including such extremely important parameters as who wrote it, and when – along with any info to help locate it. A reader/editor should not have to follow a url, and then poke around a bit, to determine whether the author is a world famous expert, or an unknown clown.
Second: all of that info – the "full bibliographic details" – applies to the source (or "work") as a whole, and need be, and indeed, should be, cited only once per source, per article. And that constitutes the full citation.
Now your question really has two parts: 1) how to cite a source (any kind) more than once in an article (i.e., "re-use" it)? And: 2) how to cite a specific location in an e-book?
To cite a source more than once, many editors use "named refs", but that is quite unsatisfactory, and forces use of such crudities as "rp". A much better practice is the use of short citations (short cites), such as "Smith, 2001", that refer back to the full citation. These can be done as straight text, or with various templates, such as harvnb. And you can have as many as you need, where you need them, and each one can have a specific "in-source" attribution (page number, or whatever).
As to the second question, I think you're there already: specific URLs. Just do something like "Smith, 2001, <specific url>", with whatever information you think would be helpful. Not hard at all. ~ J. Johnson (JJ) (talk) 20:35, 27 April 2017 (UTC)
Short citations are way worse than {{rp}}, and are rarely used on Wikipedia. Bright☀ 13:41, 28 April 2017 (UTC)
How are short cites "way worse"? They are pretty much standard in much academic publishing, losing out in some scientific and medical journals only where for reasons of space they use a bastardized form of citation. On the other hand, {rp} is peculiar to WP, unfamiliar, and looks ugly as sin where it intrudes into the text (being exceedingly non-transparent).
"Rarely used" is debatable. Over 26,000 articles use {harvnb} (which is only one way doing short cites), while use of {rp} is just short of 24,000. That most editors put their citations in "named-refs" (which then forces use of {rp}) is hardly an informed choice: most editors, when they start using citations, look around, and, like puppies, do it the way they see "everyone else" doing it. And having, with great difficulty, learned how to do citation one way, they don't want to get off that horse. ~ J. Johnson (JJ) (talk) 22:50, 28 April 2017 (UTC)
There are over 5 million articles on the English Wikipedia, 0.5% is pretty rare. But let's not argue; I encourage you to add or fix citations on articles with short citation style. Compare <ref name="author-year" />{{Rp|pages}} with <ref>{{Harvnb|author1|author2|year|pages}}</ref>. Look at all the redundant info, and if you add or change the name of one of the authors, or if there's no author name, or if you change another minor detail in the ref, the short reference breaks, so maintenance is needlessly complex. With named references, you change the reference once and nothing breaks. And of course, the reference section itself is much more concise with long refs than short refs.
In bullet points:
  • No duplication in the reference section (short refs require two separate ref sections)
  • Only one click required to get to the complete reference (short refs require two)
  • Direct links to and from a reference and each of its usages (short refs cannot be linked back)
  • Updating requires editing only a single place (short refs require updating all places where the ref occurs)
Bright☀ 19:02, 2 May 2017 (UTC)
  • Let's not wander off topic... the question was not about citation styles.. but how to precisely cite e-books (which don't have page numbers), which is problematic no matter what style one uses. Per SAYWHEREYOUGOTIT... if you got it from a kindle or some similar device, that is what you should cite. There is no URL to cite. So the best you can do is cite chapter ( and perhaps paragraph number. Blueboar (talk) 23:38, 28 April 2017 (UTC)
It's not entirely off-topic as, given some form of identifying or linking to specific in-source locations, there is the consideration of how it should be displayed. Where Fuhghettaboutit mentions URLs I am assuming he has in mind some kind of source accessible on the Internet.
As to not having page numbers: wow. That's a step backwards, page numbering being one of the principal improvements made in the development of books as we now know them. Without page numbering table of contents are just minimal outlines, and indexing is meaningless. If e-books have no such equivalent (perhaps paragraph numbering?), then the answer to the question presented is: No, one cannot "cite ... one part of an e-book", transparently or otherwise. ~ J. Johnson (JJ) (talk) 21:26, 29 April 2017 (UTC)
Most e-readers do not display page or paragraph numbers... the best you can do is manually count paragraphs from the start of the chapter. So to form a precise citation you might have to create something like:
  • "Author, Book Title, Reliable publishing co., 2016, (Kindle edition) Chapter 3, Paragraph 527."
It might not be a standard citation format, but it would at least allow others to find the information in the book and verify that it supports what we say in our article. Blueboar (talk) 22:16, 29 April 2017 (UTC)
There is no problem citing paragraphs, but there could be problems in counting them. (E.g., is your comment just above one paragraph? Two? Or three?) Perhaps we could start a campaign to persuade publishers of such content that if they don't include paragraph numbers (like some journals do) we won't allow them to be cited. ~ J. Johnson (JJ) (talk) 23:35, 29 April 2017 (UTC)
You can also provide short quotations, using |quote= if you happen to be using citation templates. Most e-readers permit searching, so this is an easy way to find the correct part of the source. You can also use the |at= to record free-form locations (e.g., |at=dust jacket or |at=Near "The Moon, like a flower...". WhatamIdoing (talk) 06:01, 3 May 2017 (UTC)

Citation overkill proposal at WP:Citation overkill talk page

Opinions are needed on the following: Wikipedia talk:Citation overkill#Citations. A permalink for it is here. Flyer22 Reborn (talk) 06:20, 9 May 2017 (UTC)

Referencing on appearances of actors and actresses, i.e. (sometimes) filmographies

I have added sources to verify appearances of actors/actresses, like David Pevsner and Herb Mitchell (actor). The issue of verifying an actor's appearance on any television show or film has been raised, especially in Wikipedia talk:WikiProject Film. Recently, the filmography of the late Erin Moran has been discussed, while it's been removed but then reinserted. Thoughts about the filmography issue? (Frankly, filmographies had not been discussed in this project talk page [WT:V]... until now.) --George Ho (talk) 20:20, 24 April 2017 (UTC)

Standard (uncontroversial) entries in the filmographies don't need to be sourced explicitly (strictly speaking each entry is primary on its own and usually there reference sites like IMDB are given in the article which can be used for verification as well). However controversial/disputed (uncredited) appearances may require (additional) explicit sourcing.--Kmhkmh (talk) 12:54, 2 May 2017 (UTC)
With regard to the linked examples. I'd say the explicit sourcing in Presner and Mitchel is ok, but not strictly required, i.e. optional and the temporary removal at Moran was inappropriate.--Kmhkmh (talk) 13:03, 2 May 2017 (UTC)
My feeling is that uncredited appearances should be explicitly sourced or tagged/removed. (Presumably) credited appearances don't require sourcing, as the credits themselves should suffice, but if an editor challenges the (presumably) credited appearance then it's probably best to provide a source rather than debate the need for one. If an editor is consistently making such challenges it may be a matter of editor conduct. DonIago (talk) 13:00, 2 May 2017 (UTC)
Source it, and use a good source, not IMDb that is partly user-generated, but something with a semblance of proper editorial fact-checking. Wikipedia is as good as its sources. Bright☀ 18:45, 2 May 2017 (UTC)
Sorry but that seems neither consensus nor current practice and on top of it it is often not feasible. "Good" sources as in high quality (scholarly) sources simply don"t exist for most actors/actresses. Aside from current practice and practicability aspects WP doesn't really require every uncontroversial simple factoid in an article to be sourced.---Kmhkmh (talk) 20:21, 2 May 2017 (UTC)
Just because you don't have to doesn't mean you shouldn't (damn that's a lot of negatives). I've fixed plenty of articles with "uncontroversial simple factoids" that were factoids in the original sense—not true but commonly held. These factoids are as good as their sources... Bright☀ 20:58, 2 May 2017 (UTC)
As I pointed out above as long as we're talking credited roles, then every listed movie in the filmography is already a primary source for its own entry. And from my perspective this is a case of "you can doesn't mean you should" rather than " you don't have doesn't you shouldn't". From my perspective such (redundant) overcitations are not good encyclopedic writing and hence not something we should recommend. I wouldn't keep any editor from sourcing every single entry in a filmography in his article if he insists on it, but we certainly should not create the impression here that such sourcing is a recommended practice (by consensus) and I understood your first posting as the latter.--Kmhkmh (talk) 21:37, 2 May 2017 (UTC)
These "sourced-to-themselves" entries have a tendency not to check out. Bright☀ 11:13, 8 May 2017 (UTC)
I seriously doubt that a high percentage of filmography entries are incorrect, but of course mistakes (or intentional "resume" inflation) are going to happen sometimes. There's also nothing to stop someone from adding to an article a secondary source about a film that doesn't actually verify what it was supposed to. The presence of a ref in an article just tells a reader where they can verify a statement in an article. It's still up to the reader to consult that source to see if the editor who added it was correct in relying on it. So primary sources such as films are no different in that. postdlf (talk) 12:30, 8 May 2017 (UTC)
Exactly, even having a source listed doesn't guarantee the information is actually from the source. I constantly flag and remove information that's {{failed verification}}. But it's all the easier not to provide a source than to provide a bogus source, and providing a source brings the reader one step closer to verifying the information. That's why if you can source something you should source it, even if it's something as "obvious" as every single movie an actor has ever been in... Bright☀ 14:35, 8 May 2017 (UTC)

I agree that only uncredited appearances need separate/secondary sources. Otherwise the inline list entry of the film or television episode is itself the source for the credited appearance. There's nothing magical about separately formatted footnotes. postdlf (talk) 21:49, 2 May 2017 (UTC)

You could use {{cite film}} or {{cite episode}}, but in the absence of a dispute, it's probably pointless: Joe Film appeared in Joe's Film.<ref>{{cite film|title=Joe's Film|at=Credits}}</ref> does not seem very helpful. WhatamIdoing (talk) 05:50, 3 May 2017 (UTC)
The filmography of Mary Tyler Moore, which has sourcing issues, was split into Mary Tyler Moore filmography and awards. I discussed this at Talk:Mary Tyler Moore#Mary Tyler Moore filmography and awards by saying that such splitting is subject to WP:CFORK, yet the editor said that there was clutter with tables. I have wondered whether the splitting was appropriate. I can't ping the editor yet, but I may do so later. George Ho (talk) 15:23, 3 May 2017 (UTC)
I very strongly doubt that the English Wikipedia has ever seen an actual POVFORK from WP:SPLITting off the long list of awards that a famous actor has won over the years, much less from splitting off a complete filmography.
Additionally, POVFORK (and other [i.e., the many acceptable] forms of content forking) isn't really a verifiability issue, and citing the sources for such a list does not determine whether the split was a good idea. If you actually suspect a POVFORK, then you need to be thinking about WP:POV problems, not about WP:V problems. WhatamIdoing (talk) 06:52, 5 May 2017 (UTC)

Just wondering... How would this discussion affect Featured Lists? "Lauren Bacall on screen and stage" and "Ethel Barrymore on stage, screen and radio" are Featured Lists, and the appearances are cited with secondary sources. More at Category:Actress filmographies and Category:Male actor filmographies, both at Category:Actor filmographies. BTW, I found Wikipedia:WikiProject Good lists created in December 2016, the same period where lists were discussed. The Good List criteria are not yet established, however. --George Ho (talk) 07:29, 14 May 2017 (UTC)

Verified information can still be removed

@TransporterMan: You said "let's at least not misstate the policy". I agree with the sentiment, but adding that "information that is not cited... may be removed" does not match the policy. WP:BURDEN applies to information that is not verifiable, but for verifiable information, it doesn't require that the reference is in the page, only that it is known. Also if the information is not challenged nor likely to be challenged, there's no need for an inline citation. Diego (talk) 14:34, 16 May 2017 (UTC)

Not correct. As established here at this talk page again and again, uncited information may merely be removed. If there is any additional requirement on that it is only that the removing party state a "concern" that it may not be verifiable (and even that is doubtful as a requirement due to the "should" in that sentence). There is no burden on the removing editor to establish, or even attempt to establish, that the information is, in fact, unverifiable. Best regards, TransporterMan (TALK) 16:10, 16 May 2017 (UTC) Oh, and PS, the "likely to be challenged" standard only applies to the introduction of new material. The standard for removal of material after it has been added is in BURDEN and says, "Any material lacking a reliable source directly supporting it may be removed and should not be restored without an inline citation to a reliable source." — TransporterMan (TALK) 16:22, 16 May 2017 (UTC)
You forget that WP:PRESERVE is also policy, forbidding the exclusion of content that belongs in the article. The idea that any content without a direct citation may be removed for any reason, at any time, has been defended at this talk page, but it has not been encoded as such in WP:BURDEN, and is contrary to the editing policy. The "any material may be removed" that you quote has lots of qualifications and caveats in this and other policies. Diego (talk) 18:39, 16 May 2017 (UTC)
And the next question is: Does the material "belong in the article" (or not)? Nine times out of ten, that's really what is underlying most debates about a removal (and opposition to the removal)... a disagreement as to whether the unsourced material should be in the article (or not). Blueboar (talk) 19:16, 16 May 2017 (UTC)
Right, but BURDEN may only be invoked under a genuine belief that the content can't be supported by reliable sources. Deleting content merely as a way to force others to source it, without oneself checking whether it's easy to verify, has been deemed disruptive by previous consensus. Diego (talk) 21:48, 16 May 2017 (UTC)
That's absolutely false and has been refuted here again and again. Point to that consensus if you can. Moreover, PRESERVE is expressly conditioned upon V being satisfied before PRESERVE kicks in: "as long as any of the facts or ideas added to an article would belong in the 'finished' article, they should be retained if they meet the three article content retention policies: Neutral point of view (which does not mean no point of view), Verifiability and No original research" and "Several of our core policies discuss situations when it might be more appropriate to remove information from an article rather than preserve it. Wikipedia:Verifiability discusses handling unsourced and contentious material" (emphasis added). (And that begs the question of whether PRESERVE, though designated a policy, really requires anything at all.) Best regards, TransporterMan (TALK) 22:48, 16 May 2017 (UTC)
Once upon a time, the policy used to say "However, this policy should not be used to cause disruption by removing material for which reliable sources could easily be found — except in the case of contentious material about living persons, which must be removed immediately." I still feel that this should be true to some extent. But, just so we're clear: One should not go around deleting relevant material that meets the other guidelines if sources can easily be found, should one? Sławomir Biały (talk) 23:31, 16 May 2017 (UTC)
That's already provided for in the policy: The burden for seeking out sources, whether easy or difficult to find, should be on the editor who initially adds the material or who restores it after it has been deleted, that is, the persons asserting that the material should be in the article. The burden should not be on the person who removes unsourced material. (Just as it should not be on a person who challenges unsourced material without removing it.) Per the policy, the only burden on the remover or challenger should be (if any, see above) that they have and state a concern that the material is unverifiable. Best regards, TransporterMan (TALK) 04:05, 17 May 2017 (UTC)
Thanks, Blueboar. Your new version is much closer to the wording at WP:CHALLENGE. Diego (talk) 14:52, 16 May 2017 (UTC)

I think that the February 2017 wording was better, but am ready to compromise to really settle it. But not open to creating a shifted starting point for future discussions. So I'm not sure whether to revert or not. North8000 (talk) 15:38, 16 May 2017 (UTC)

We could rather revert to the pre-february 2017 version as a starting point. It was a long-standing stable version, after all. Diego (talk) 15:51, 16 May 2017 (UTC)
According to the RfC there does not seem to be a consensus to change from pre-february 2017 version, so why have the changes not been reverted back to that version? -- PBS (talk) 15:59, 16 May 2017 (UTC)
FYI... No problems on my end with a revert to the pre Feb version. Blueboar (talk) 16:09, 16 May 2017 (UTC)
I'm also fine with the pre-Feb version, but I'm not sure that there's any real consensus for anything in the foregoing discussion, at least not a consensus of the quality needed to change this core policy. Best regards, TransporterMan (TALK) 16:11, 16 May 2017 (UTC) And PS, I've reverted to the post-Feb version as the last at least semi-stable version. That does not mean that I support that version since I actually support the pre-Feb version, but if we're going to continue to draft on this it ought to be done here on the talk page, not in the policy. Best regards, TransporterMan (TALK) 16:40, 16 May 2017 (UTC) And PPS since there's a continued discussion about closing, above, I think I need to make my position clear about my preferences. My preferences are, in order from most preferred to less preferred, (a) The February text with the pre-Feb heading (I think I'm the only one asserting that combination), (b) The pre-Feb text and heading, (c) The Feb text and heading, and (d) any of the post-Feb substitutes which have so far been proposed (here or by editing the article). I find any of (a) through (c) acceptable in the order of preference just stated, but I do not find any of the (d) versions acceptable and object to the adoption of any of them. Best regards, TransporterMan (TALK) 04:14, 17 May 2017 (UTC)

The Feb version was arrived at in a 27,000+ word discussion that ran for 23 days (Feb 7th - March 2nd) with no objections at the end of the process. And it merely summarized the actual policy. I'm ready for compromise be we need to recognize that. North8000 (talk) 18:38, 16 May 2017 (UTC)

I think that we are adding a layer to the core difference of opinion that unnecessarily moves it from challenging to hopelessly mired. The core difference of of opinion, is, beyond / stipulating meeting verifiability being a requirement for inclusion, to what extent is or isn't it a force or reason for inclusion? The extra layer that is miring it down is that the discussed wording wades into a bunch of other areas and procedures that are not even in dispute. For example , rules for removing material. North8000 (talk) 23:54, 16 May 2017 (UTC)

  • Comment -- I strongly disagree with the statement that WP:BURDEN applies to information that is not verifiable. This is very vague -- who determines whether information is "verifiable"? Similar to TransporterMan, I don't believe that consensus for such an interpretation exists. K.e.coffman (talk) 02:18, 17 May 2017 (UTC)
    • Perhaps it will make more sense like this:
      1. If you personally know that a given piece of information is verifiable, then you personally should not remove it solely on the grounds that nobody has yet typed a citation after that information.
        You could add sources, you could tag it, you could move it to another article, you could copyedit it, you could do any of the other suggestions at WP:PRESERVE, you could even remove it on the grounds that it's WP:TRIVIA or WP:POVish or WP:SYNTH, but you personally should not remove information that you know is verifiable with a claim that you think it is not possible to verify it, when you have actual, positive knowledge that it could be verified. Put more simply, if you cannot comply with BURDEN's request to "please state your concern that it may not be possible to find a published reliable source for the content" without telling lies, then you should not remove that material under BURDEN.
      2. However: if you personally do not already know that the material is verifiable, then, yes: You may remove that material. There are no other requirements: if it is uncited, and if you personally do not already know that the material is verifiable, then you are entitled to blank it. I would only add to this that you please remember that it is both uncollegial and anti-policy to take actions that you personally believe will have a (net) negative effect on an article.
    • And: Remember that every editor knows different things. So while I couldn't ethically claim BURDEN for, say, the recent news about genetic treatment for sickle cell disease, most editors probably could. "I can't claim BURDEN" is different from "Nobody can claim BURDEN". WhatamIdoing (talk) 16:37, 17 May 2017 (UTC)
      • Re: "If you personally know..." -- How would one know what's in another editor's head? The current interpretation is cleaner, IMO. K.e.coffman (talk) 18:08, 17 May 2017 (UTC)
        • You don't need to know what's in another editor's head. You only need to know what's in your own head. I would be acting wrongly if I were to remove uncited material that I personally know to be verifiable, and then claim that I am doubtful that it is possible to verify this. The policy does not authorize me to tell deliberate lies about the potential verifiability of a statement, even if I feel certain that I won't get caught. Another editor, looking at the same sentence, but without (for example) the benefit of having read the same news article yesterday, could invoke BURDEN quite honestly for the same material. As Diego succinctly put it, "BURDEN may only be invoked under a genuine belief that the content can't be supported by reliable sources". Your genuine belief in the material's non-verifiability may be genuinely and honestly mistaken, and your genuine belief is allowed to be based on nothing more than a wild guess, but you must not blank uncited material under BURDEN when you have actual, positive knowledge that the material is verifiable. So the policy won't support me blanking uncited material about the genetic treatment of sickle cell that was reported this week, but (perhaps) you could; on the other hand, I could lazily challenge the verifiability of any number of uncited sentences in WWII-related articles, and you, by virtue of your greater knowledge about WWII, would be restricted from making the same claims. WhatamIdoing (talk) 05:33, 19 May 2017 (UTC)
One of the most difficult things for editors who are knowledgeable about a subject to accept is that their knowledge can be challenged by editors who are not knowledgeable. Yet this is true... unsourced information can be challenged (and removed) on the mere belief that it is unverifiable. It is up to those who know the information is verifiable to provide citations when citations are requested. Blueboar (talk) 13:13, 21 May 2017 (UTC)
I think that a main point of WhatamIdoing's post is that you should not do a wp:ver-based removal unless you have a sincere question about its verifiability. I wish that that point were more prominently enshrined in policy. North8000 (talk) 13:38, 21 May 2017 (UTC)
Yes... but but she goes on to say that a sincere question may be based on a good faith belief of unverifiability. That is an important point for those with knowledge of verifiability to remember. Others don't necessarily know what you know, so it is up to you to supply the sources when sources are requested. Blueboar (talk) 14:01, 21 May 2017 (UTC)

How to describe the Emmett Till case in the lead sentence of the Emmett Till article

Opinions are needed on the following matter: Talk:Emmett Till#RfC: Should we include the "accused of showing an interest in a white woman" aspect in the lead or specifically the lead sentence?. A WP:Permalink for it is here. Flyer22 Reborn (talk) 17:07, 24 May 2017 (UTC)

Encouraging accessibility

All else being equal, a more accessible ref is better than a less accessible one. For instance, if there are 3 copies an article: behind a paywall, behind a loginwall, and freely accessible online, the latter is the better source: easier for both readers and verifiers to access.

Can we rephrase the 'Accessibility' section to call this out? While access isn't a requirement for a source, the more readers that can access it themselves, the more effectively a claim can be verified. – SJ + 08:45, 25 May 2017 (UTC)

Current

Some reliable sources may not be easily accessible. For example, an online source may require payment, and a print-only source may be available only in university libraries. Do not reject reliable sources just because they are difficult or costly to access. If you have trouble accessing a source, others may be able to do so on your behalf (see WikiProject Resource Exchange).

 
Suggested

Some reliable sources may not be easily accessible. For example, an online source may require payment, and a print-only source may be available only in university libraries. Do not reject reliable sources just because they are difficult or costly to access. If you have trouble accessing a source, others may be able to do so on your behalf (see WikiProject Resource Exchange).

That said, all else being equal, a source freely available to read online is preferable because more readers will be able to verify its claims. If two sources are equally suitable to verify a claim, accessibility is a reason to prefer one over the other.

  • Oppose. The proposed added content is apparently in conflict with the existing policy. The key words here are "all else being equal". If all else were truly equal, then this is an obvious point: if the same source can be accessed by a publicly available link, then, subject to policies that allow such links, of course we can give a link to the source since this makes it easier to verify the source. I don't think this needs to be articulated in policy. But I believe that the proposed addition will do more harm than good. Firstly, all else is usually not equal. Two sources may be adequate to verify the same claim, but one of them is a better source. For example, a standard academic textbook is probably the best source for scientific types of information, despite not necessarily being publicly accessible. Some things could probably also be verified by linking to Professor X's online lecture notes, but that is very likely to be an inferior source despite being more accessible. I am strongly opposed to any changes to the policy which suggest that sources of this latter kind are in any way better than sources of the former. Too much here hinges on the interpretation of "all else being equal". Secondly, there are already policies against linking to certain sources (WP:LINKVIO). Sławomir Biały (talk) 10:25, 25 May 2017 (UTC)
This strikes me as a sensible change and not a radical departure from existing policy at all. I think the key focus is on equally suitable to verify a claim. A clear example is two textbooks, one of which is closed access and the other open access, but both respected. Or a situation in which there are two newspaper articles, one of which requires a subscription and the other does not. If they are both mainstream newspapers with similar reputations for fact-checking and accuracy, then all else being equal (which doesn't really mean "identical"), the more accessible one is generally preferable. It's not a hard rule--a "must do"--but a small nudge towards what will benefit more readers. Cheers, Jake Ocaasi t | c 11:09, 25 May 2017 (UTC)
@Sławomir Biały:, I don't think anyone would disagree with your post, but I think you are misinterpreting "all else being equal". This implies that both sources A, B, and C are equal. Substituting lecture notes for a textbook doesn't fall under "all things being equal" in the proposed claims so I don't see any issue with him using those choice of words. In my mind, any clarification like this is pretty minor, but at of the day ultimately helpful. I am in support. SEMMENDINGER (talk) 11:11, 25 May 2017 (UTC)
If "all else being equal" is really open to this kind of misinterpretation, then it is clearly unacceptable. Sławomir Biały (talk) 12:16, 25 May 2017 (UTC)
I agree with Sławomir Biały. It's obvious that if exactly the same source is available online for free, and also through a paywall or in paper form, a link to the free version should be provided (possibly in addition to the information on the paywall/paper version). I don't think that needs to be spelled out. What isn't obvious is how editors will interpret "all else being equal". All else will hardly ever be equal, in terms of the reputation of the author and publisher, and the ability of the citation to help the reader find other related sources. My fear is editors will interpret "all else being equal" with the much weaker criterion of both sources supporting the claim. Jc3s5h (talk) 11:16, 25 May 2017 (UTC)
The overall change still makes sense to me but these objections suggest a rewrite of the proposed addition and removing the "all else being equal". How about simply: That said, more readers will be able to verify a source freely available online. If two sources are equally suitable to verify a claim, accessibility is a reason to prefer one over the other. Cheers, Ocaasi t | c 11:19, 25 May 2017 (UTC)
(EC) I think the "all else being equal" part is confusing, so I'd suggest to shorten the addition to basically just the second one of Sj's sentences, i.e. something like

That said, if two sources are equally suitable to verify a claim, accessibility is a reason to prefer one over the other.

Ocaasi's phrasing just above would also be fine to me. -- Daniel Mietchen (talk) 11:21, 25 May 2017 (UTC)
Seems impossible to judge without examining the two sources side by side, so I am not a fan. Although I think I have no problem with saying something like, 'if the exact same source is located in different places, absent copyvio problems at the link, link to a freely available online source is preferred (while stability for a future possible dead link should also be considered).' Alanscottwalker (talk) 11:23, 25 May 2017 (UTC)
  • (edit conflict) This is kindof undone by the very problem it tries to solve: If two editors disagree over which source to use, because one of them is in print or behind an paywall, presumably one of them probably doesn't have access to the source in order to evaluate whether all things actually are equal. So this effectively becomes a "prefer online sources" clause. However, I would totally support something purely inclusive rather than potentially exclusive, for example When the same basic information is also freely available online, consider using both sources to support the content, to maximize the verifiability for the largest number of readers. TimothyJosephWood 11:33, 25 May 2017 (UTC)
I worry very strongly about the potential for abuse. Doesn't this encourage the citation of open access papers? This should be avoided. Sławomir Biały (talk) 12:16, 25 May 2017 (UTC)
Umm... I suppose we could say freely available in reliable sources online if it's not completely clear that all the constraints of WP:RS are still implicit. And if someone is blatantly ref spamming, that's a countervandalism issue and fairly independent of good faith editing generally.
I imagine this being most relevant in older or more esoteric articles, especially those that use the old footnote/bibliography citation style and rely heavily or entirely on print sources. In those types of situations, we're more-or-less clarifying that This is an acceptable way of writing, but if it's possible to accompany this with accessible online sources, it is an objective improvement in the article and a worthwhile endeavor. TimothyJosephWood 12:43, 25 May 2017 (UTC)
TJW: +1 :) – SJ + 12:51, 25 May 2017 (UTC)
If we're simply clarifying a best practice, then the appropriate place to do it would be in an essay, not in the main verifiability policy. I stand by my strong conviction that we should not be adding anything to this policy that apparently encourages linking to open access literature. Furthermore, experience tells me that the average Wikipedia editor (which probably includes none of the present company) cannot tell the difference between an authoritative dead tree source and a substantially less suitable online source. So we absolutely should not encourage substituting one for the other. In contrast, I do think editors are already more than capable of deciding for themselves whether including a link to an online source is an objective improvement to the article, and that this should be assessed on a case-by-case basis rather than as a blanket policy-level rule. Sławomir Biały (talk) 13:06, 25 May 2017 (UTC)
"Instruction creep is bad" is one argument entirely. But "case by case basis" implies that there would be some fairly common instance where the inclusion of a quality freely accessible source actually wouldn't be an objective improvement. I'm having problems thinking of one that a qualifier like "quality" or "reliable" doesn't completely eliminate. TimothyJosephWood 13:17, 25 May 2017 (UTC)

While I generally agree with the general idea, I'm not convinced that this needs to be legislated explicitly and also I worry as well about abuse potential.

Some editors might feel encouraged to systematically replace offline sources by online sources of supposed equal quality and as often with such ventures do it in rather formal at many articles where they aren't really familiar with the topic or its details.

Another problem is that many online sources might get offline over time and even vanish completely and in that case switching to currently easily available online source may be deterioration in the long run and lead to unsourced content.

Yet another issue might be with what we consider "online". Is it supposed to freely accessible material only? Or does it include online sources behind paywalls? For the latter the assumed advantage of many readers being able to verify the content is significantly lower.--Kmhkmh (talk) 18:48, 25 May 2017 (UTC)

P.S. To add another thought, rather than replacing an offline source but one could simply add a reliable online source as a second source. Ideally editors should verify some content against several sources anyhow and while there is usually no need to cite a second reliable source, there isn't really anything blocking them from adding an additional online source, when appropriate and convenient to readers.--Kmhkmh (talk) 18:58, 25 May 2017 (UTC)

Take II

Ok, let's avoid confusing language. I see two good variants suggested:

When information from a closed source is available from a freely accessible source, consider also referencing the accessible source, to enable verification by the largest number of readers. See dedicated section below.
When two sources are highly similar and equally suitable to verify a claim (e.g., multiple versions of the same document), accessibility is a reason to prefer one over the other. [a reason, not a requirement; to limit redundancy]

Thoughts? – SJ + 12:51, 25 May 2017 (UTC)

(I do think the secondclarification is worth stating, since publishers tend to have style guidelines that privilege citing closed versions of documents, and hide the existence of open versions. This benefits the publisher, but not the reader. – SJ +)
In #1, I would add "also referencing" to head off petty edit wars over the interpretation of "prefer" (i.e., prefer means we use my source and remove yours). Again, thinking mostly about books vs. websites, and less about peer reviewed journals. In these cases, books are going to usually be the higher quality but less accessible source, written by better qualified authors, who put more time into the publication, but the basic information may likely still be online, even if it may have overall less depth or nuance. TimothyJosephWood 13:03, 25 May 2017 (UTC)
Added. Agreed that there is a bimodal use case here: the typical case w/ books and w/ journals feel quite different. – SJ + 15:44, 25 May 2017 (UTC)
I can support saying that it is appropriate to "improve" the accessibility of an existing citation, by adding a link to an on-line copy of the same cited source. I can not support favoring an on-line source over a different (and potentially better) non-on-line source. Blueboar (talk) 13:16, 25 May 2017 (UTC)
I don't see an obvious reason why it would need to be the same source, rather than just some source which is reasonably reliable and confirms the article's content. TimothyJosephWood 13:46, 25 May 2017 (UTC)
Timothyjosephwood's formulation "some source which is reasonably reliable and confirms the article's content" isn't good enough. A free source where the author, publisher, or both has a lesser reputation than a non-free is inferior. It could be an additional cited source, but should not be cited instead of the non-free source. Another reason to consider the free source to be inferior is if the free source is in any way preliminary, tentative, etc. while the non-free source is final, authoritative, official, etc. Also, a free source that is merely an excerpt or summary of a more complete non-free source is inferior. Jc3s5h (talk) 14:47, 25 May 2017 (UTC)
I'm personally arguing for inclusion in addition to rather than in place of. If a source is inferior in some way, but still generally reliable, and overall more easily accessible, and it is in agreement with the other superior but inaccessible source, then it should probably be added in nearly every case. TimothyJosephWood 14:57, 25 May 2017 (UTC)
We want to cite the best sources possible for any information we present. It may be that a less accessible source gives a more complete explanation of the information than an easily accessible source... an off-line source may present the information in better context than the on-line source... it may cite additional sources (sources that the on-line source does not cite). There are lots of reasons why we might favor a less accessible source over an easily accessible one. On-line accessibility is nice, but the quality and completeness of the source is far more important. Blueboar (talk) 15:12, 25 May 2017 (UTC)
That's... still not really an argument against including both when available as a matter of policy. TimothyJosephWood 15:23, 25 May 2017 (UTC)
Including both is nice... but not necessary Blueboar (talk) 16:31, 25 May 2017 (UTC)

Update: I've separated out the sentence encouraging adding accessible sources where possible, in the section below. – SJ + 08:16, 26 May 2017 (UTC)


  • Oppose Adopting this proposal will increase bias for online works and also bias against higher quality non-free sources. The normal thing to do in libraries and research is to direct everyone to seek out the highest quality sources which they can access. I think Wikipedia should follow that precedent without complicating the situation with other goals which could compete against quality. I have am not aware of experiencing or hearing that Wikipedia editors routinely encounter a free/non-free decision, so I think that adopting this would more often risk or cause a drop in referencing quality than it would exchange non-free sources for equally suitable free ones.
If there are citations to a high-quality non-free source, then I think additional citations to even a lower-quality free online source are a great supplement to that. I think in most cases, just adding another citation is preferable to removing something or putting the work into a conversation. Blue Rasberry (talk) 16:52, 25 May 2017 (UTC)
  • Oppose. I use many sources that are not available for others to read. They are often the best sources for the content. QuackGuru (talk) 16:54, 25 May 2017 (UTC)
  • Yeah, I oppose this too. I do think that we should stress, just please, please, please relate the very best reaserch sources possible, and not muddy it otherwise. Alanscottwalker (talk) 17:25, 25 May 2017 (UTC)
  • Oppose This seems way worse as the original suggestion.--Kmhkmh (talk) 18:50, 25 May 2017 (UTC)
    What aspect is worse? – SJ + 08:16, 26 May 2017 (UTC)
  • Oppose. Use the best sources. Whether a source is freely accessible or not is not, and never has been, a factor in determining the reliability of a source (WP:RS). There are already many instances of new media and open access publications, where being available electronically at the exclusion of print media sources is already a red flag regarding the reliability of a particular source. (In fact, this is a continuing problem that we routinely have to deal with.) I would just as soon not have the opposite enshrined in our verifiability policy. There is nothing in the current policy that would forbid including links to sources, or freely accessible sources. So it is already the case that, if these happened to be the best sources anyway, they would already be included. So the proposed addition (either version) appears to solve a non-existent problem. Sławomir Biały (talk) 19:27, 25 May 2017 (UTC)
  • Oppose. The primary basis, perhaps even the sole basis, for preferring one "source" over another is its authority (which gets wrapped up into what we call "reliability"), not its convenience of access to WP editors. Nor is there any need to "prefer one over the other", unless one imposes some kind of limit on the number of sources. If an editor has access to an authoritative source which is otherwise inaccessible, but other sources exist which describe the point (likely citing the authoritative source), then it is perfectly sensible to cite ALL relevant sources. The proposal rather implies that it is a matter of one source OR another, but that is false. Use all the sources that are useful. Only if two sources are entirely "equally suitable to verify a claim" (perhaps multiple eye-witness accounts?) should other considerations come into play, and then it should be fine to leave it to the editor's discretion. ~ J. Johnson (JJ) (talk) 20:55, 25 May 2017 (UTC)
    I agree with most of this. Caveats: We do sometimes limit the total # of sources (if you simply double the number of sources on a given page, that would annoy both editors and readers). Convenience & speed of access to other editors improves confidence in the appropriateness of the cite. And there are regular cases where multiple sources have entirely comparable authority (two national newspapers repeating the same 3-paragraph summary of a current event; two websites hosting a PDF of an article).
  • Oppose. The quality of the source and whether it's appropriate for the point in question are the only two considerations. Of course if a source is available that is good enough and is free, it can be added too. There's no need to choose. This is the kind of thing that can be added to WP:RS, rather than to the policy. SarahSV (talk) 21:17, 25 May 2017 (UTC)
    There's definitely a piece here related to RS / cite-style ("please don't remove a reference as 'redundant' when it is more accessible than others for a statement"), and a piece related to V ("the harder it is to access a source, the harder it is for readers + editors to use it to verify a statement"). – SJ + 08:16, 26 May 2017 (UTC)
  • Oppose original proposal and Take II. While I appreciate the idea here, without a long list of do's and don'ts appended (adding a burden, limiting to 1RR, others), this is just a drama magnet and another requirement for POV-pushers to wikilawyer to throw up barriers to their domination of articles, with no benefit to the encyclopedia which outweighs that disruption. — TransporterMan (TALK) 21:23, 25 May 2017 (UTC)
  • Counter proposal: When the same source is available from multiple locations, the location with the greatest ease of access should be used by default. Jclemens (talk) 21:41, 25 May 2017 (UTC)
Could run afoul of our WP:SAYWHEREYOUGOTIT policy... if you found the information in a hard copy version (say, in a library) you should cite that... not the on-line version you didn't read. Blueboar (talk) 22:18, 25 May 2017 (UTC)
Would we need to define 'same'? If I read the print version, and cite the online version of the same article, that *is* saying where I got it. Jclemens (talk) 23:36, 25 May 2017 (UTC)
@Blueboar: This seems to be a bit of an artificial problem. If there is indeed and online version of the same source available, then there should be no problem to read/check that as well. After all it is not "say where you first got it".--Kmhkmh (talk) 01:30, 26 May 2017 (UTC)
  • Jclemens, can you think of a practical example where adding that to the policy would result in something useful that doesn't happen already? In some cases, linking to the free version instead is not a good idea. For example, academics may retain the copyright of their work and host it on their own websites, or obtain permission to do so from the publisher. But those personal links degrade faster than the publishers' links or the jstor or doi link. Including the free in addition to the others is a good idea, but not instead of and not as the default. SarahSV (talk) 22:18, 25 May 2017 (UTC)
    • No, not really. But if an article is cited correctly and thoroughly, how difficult is it, really, to fall back from one freely hosted version to the paid version if and when the free version goes away? I think of providing the most available/accessible source as a user conduct issue: If an editor prefers the inaccessible, offline version... why? Jclemens (talk) 23:36, 25 May 2017 (UTC)
  • As I said, the less accessible (not necessarily offline) version might be more stable. That's why I'd like to see some practical examples from the people making these proposals. We can add free links subject to copyright and other policies. I can't think of any sourcing situation where we're allowed to add one link only. Some examples are needed to show that this is a real issue. SarahSV (talk) 23:42, 25 May 2017 (UTC)
  • Stability is important. Journals are not archival: URLs can change, and if the journal closes its website may also. In terms of stability, personal website < journal < university library/repository (the latter are quite robust and often tied into archival mirrors like LOCKSS). Massive dedicated repos such as JSTOR are a hybrid, comparable to university repos. – SJ + 08:16, 26 May 2017 (UTC)

"Ideally editors should verify some content against several sources"; "use many sources that are not available for others to read. They are often the best sources for the content."; "not its convenience of access to WP editors"; I think these three quotes (I could have included more it is just those caught my eye on first reading the comments), show a lack of focus on why citations are included. Citations are not included for the convince of editors they are there to verify allow readers to have confidence that the WP material they read is accurate. This encyclopaedia is written with the intention that the majority of it will be read online. Of course editors should verify content against several sources, yes they may have to use sources that others can not easily access, and citations are not included for the convenience of access to WP editors; However articles are written for the convenience of WP readers (it is part of the raison d'être for its existence). So I think that WP editors ought to consider that when deciding on the material to be cited, and this policy ought to remind editors of why citations are included. -- PBS (talk) 06:13, 26 May 2017 (UTC) @PBS:

I think you got that partially wrong.
WP content is designed to be read online and its sources are not necessarily. Furthermore the selection of sources is primarily guided by quality (reputation, reliability) rather than ease of access and the service to readers is a highly reliable/reputable source confirming the content rather than just some online source claiming the same.
Aside from that one should keep in mind that editors are readers as well and particularly of the reader type that checks sources, whereas the majority of readers doesn't.
Another thing to keep in mind is, that the sourcing requirement is primarily a quality assurance tool to compensate for the lack of central editorial control only allowing qualified experts as editors. Because WP cannot guarantee the authoritative domain knowledge of its editors it uses the authoritative domain knowledge of its sources as a replacement. Aside from pointing out that sources are more a tool than product this perspective emphasizes as well that the reputation/authoritativeness of a source is more important than its online accessibility.--Kmhkmh (talk) 16:01, 26 May 2017 (UTC)
Well put, especially regarding quality assurance and authoritativeness. My criticism of PBS' points is this: the expectation that "editors should verify content against several sources" is incomplete. Ideally they should survey the topic as broadly as possible, using authoritative primary and secondary sources to determine what the most authoritative sources are. Often this requires critical assessment, as expert opinion may change over time (e.g., new discoveries), or be conflicted. Ideally the editor becomes expert enough to identify the key sources, and then cites those. Which may be "several", but not simply the first several sources found on the Web. ~ J. Johnson (JJ) (talk) 21:20, 27 May 2017 (UTC)

Twitter

I've never seen Twitter used as a source in any other article other than the Laurel Van Ness article. I don't know if Twitter is good for sourcing information. Has there been a discussion and was a consensus reached? Mr. C.C.Hey yo!I didn't do it! 07:49, 20 May 2017 (UTC)

Yes... we have had many discussions about twitter... and the consensus is that twitter posts are not reliable. First, since there is no fact checking on twitter, any tweets would (at best) only be verifiable for the fact that person X tweeted "Y", but not for the accuracy of "Y". Second, even then, it is too easy to spoof a twitter account. This means we can not always be sure that tweets are actually written by the person we think they were written by... so they are questionable even as to the tweeter's opinion. Finally, twitter posts are ephemeral... they are automatically deleted after a certain time... and thus will not be verifiable once they are deleted. Taken all together, these issues make twitter unreliable.
That said, in today's world, newspapers do report on the tweets of notable people (such as President Trump)... if a tweet is reported on by a newspaper, we can assume that the tweet was made, and that the newspaper has verified that it came from the notable person. At that point we can report that person X tweeted "Y", and cite the newspaper for this fact... but we still can not say that "Y" is accurate. And finally... Note that even this may not be enough for us to mention the tweet... venerability does not guarantee inclusion... there are other policies (such as WP:NPOV) that might cause us to omit mentioning the tweet. Blueboar (talk) 12:41, 20 May 2017 (UTC)
Before I saw this response, I seen another article that used Twitter as a source. I needed a response before removing them. Thanks. Mr. C.C.Hey yo!I didn't do it! 07:07, 21 May 2017 (UTC)
I disagree a bit here as this really context dependent. Meaning all the information that might be sourced with an official webpage might as well be sourced with any official social media account (including twitter). Consider for instance uncontroversial/uncritical biographical information, that is taken from a person's website.--Kmhkmh (talk) 19:04, 25 May 2017 (UTC)
All sources are reliable for something, and all sources are unreliable for most things. Context matters. "Donald Trump tweeted that he ate a cheese sandwich" is verifiable to a tweet from his verified account that describes his meal. "The King of Ruritania is a commie mutant traitor" is not something you could verify to a tweet. WhatamIdoing (talk) 07:45, 28 May 2017 (UTC)

Proposal: "consider adding accessible sources"

Separating this idea from the section above. Proposal: to add the following line.

When information from a closed source is available from a freely accessible source, consider also referencing the accessible source, to enable verification by the largest number of readers.
  • Comment: Echoing PBS above, I see verifiability as important for two reasons: as a standard for knowledge included in articles, and as an affordance for readers to check for themselves that this standard is being met. Inaccessible sources are bad at the second point, and make it harder for the community to confirm the first point. The more readers are able to access the source (accessibility / legibility), and the more quickly they are able to verify the claim (precision of sourcing), the more meaningful verifiability is. – SJ + 08:16, 26 May 2017 (UTC)
The problem I have with all these proposals is that they seem to be equating "accessible" with "on-line"... but a source does not need to be on-line to be accessible. A hard copy book that is in most public libraries, or that can be purchased for a few dollars on amazon is just as accessible as an on-line source (although not as immediately" accessible)
As to this version, I hope you realize that "consider" means we can ignore whatever follows (since we can always say: "Ok, I've considered it... and choose not to do it.") Blueboar (talk) 11:05, 26 May 2017 (UTC)
That'll never fly. Once you start with "consider this", people will read it as "you should", and from there it's a very small step to "required". Although it's not a common belief (and less so since I re-wrote WP:POLICY to contradict it), we have seen editors profess their sincere belief that "you should" in a policy indicates a near-absolute requirement, on the grounds that an IAR-based philosophy prevents the community from writing "you must" in a policy when we actually mean "you must". WhatamIdoing (talk) 08:03, 28 May 2017 (UTC)
  • Comment: It seems this encourages links to rather crappy sources. One supposes it's possible for a partisan op-ed to have the same bit of "information" as an academic journal article, for some unknown definitions of same information (just think of the wonderful debates about same information). Given context matters, and that say, promoting links to partisan op-eds is not the reason for verfiability, this seems like a poor idea. -- Alanscottwalker (talk) 11:11, 26 May 2017 (UTC)
I share Alanscottwalker's concern. The scenario is first a good summary of an excellent print source is added. Then someone comes along and adds an inferior free online source that supports the claim. Then an editor comes along, reads the free source but not the print source, and revises the claim so now it agrees with the inferior source but not the superior source. Jc3s5h (talk) 13:05, 26 May 2017 (UTC)
I disagree with "consider also referencing the accessible source". Too many citations at the end of a sentence decreases readability. One citation is enough for each sentence in most cases. QuackGuru (talk) 14:25, 26 May 2017 (UTC)
I agree with this -- PBS (talk) 10:11, 27 May 2017 (UTC)
  • Oppose that still has the same problems pointed out above not to mention that it doesn't really extend but contradicts current policy, as the quality/reputation of a source doesn't seem to matter anymore.--Kmhkmh (talk) 16:04, 26 May 2017 (UTC)
  • Oppose as before. While most of the regulars here at V recognize that the "consider" makes this optional, others will incorrectly assert it as required. This whole topic would make a good essay, but doesn't need to be in policy or guidelines. Regards, TransporterMan (TALK) 16:59, 26 May 2017 (UTC)

What some are writing, puts me in mind of the Proclaimers song "I would walk 500 miles". While I don't think the wording suggested so far is suitable, I think it would help if people could first agree (or disagree) that ease of access is desirable "when [the source is] of equal quality and relevance."

This is already touched upon in the section "Accessibility", both in the sentence "Do not reject reliable sources just because they are difficult or costly to access" implies that most editors consider ease of access to be desirable (if not, then this sentence would be unnecessary); and "However, because this project is in English, English-language sources are preferred over non-English ones when available and of equal quality and relevance." which if ease of access was not considered desirable this sentence would not be needed.

One of the things I have repeatedly defended over the years in WP:PSTS is the wording that currently says "reputably published". This is included to stop editors using unpublished primary sources to include facts that are not in any secondary sources (OR). However it has a secondary affect making sure that the cited information can be found relatively easily by others.

@Blueboar and Alanscottwalker I have some books that are short run first editions. If I use these books for citations then they tick all the boxes as a reliable source, but they are hard to find (or could be very expensive to purchase, if bought just to verify a fact in an article). There maybe later editions (and so different pagination) that are much easier to access (common edition in many public libraries), or the information may also be accessible through access to the pages online on the Google book website. Which of the three sources would be the "best" (most useful) in include for readers to verify a fact in a Wikipedia article?

-- PBS (talk) 10:11, 27 May 2017 (UTC)

Well, in my very first comment on this topic I said, Seems impossible to judge without examining the two sources side by side, so I am not a fan. Although I think I have no problem with saying something like, 'if the exact same source is located in different places, absent copyvio problems at the link, link to a freely available online source is preferred (while stability for a future possible dead link should also be considered).' Alanscottwalker (talk) 11:23, 25 May 2017 (UTC)
I think, Jc3s5h illustrates well a problem with the current proposal. Given that for V, "a citation to a reliable source that directly supports the contribution" (emphasis added) is required - the complete fungibility of contribution wording and source wording cannot be taken for granted. And given WP:CONTEXTMATTERS, we are in an uncertain territory of trying to predict the precise relationship between contribution and source, and that's even before we get to the contribution having to be NPOV in presentation and not OR in context. (As for google book links, my understanding is that they change what is available to see, in part to avoid copyright problems - so yeah it's probably good for awhile). But like I said, I think I have no problem with adding a link in the source citation when it appropriately exists. Alanscottwalker (talk) 11:06, 27 May 2017 (UTC)
  • Oppose: Many editors go out of their way to cite the sources that are the best ones available, as they should, and do so consistently. Besides, we cite sources to answer the question WP:SAYWHEREYOUGOTIT, not where else this information can probably found. Faithfully following the proposed policy would lead to extremely cluttered articles that for, say cited every page of an academic book, also cite whatever website that happens to contain a passing mention that happens to verify the cited fact as well. – Finnusertop (talkcontribs) 16:57, 27 May 2017 (UTC)
  • Oppose -- pointless and not in the spirit of using the best sources available. K.e.coffman (talk) 19:50, 27 May 2017 (UTC)
    • We don't actually require, or even explicitly encourage the use of "the best sources available". (If we did, then we'd have to agree on what makes a source "the best": Academic-ness? Accessibility? Likelihood that the citation will be useful in 10 years? Are books "better" than newspapers? WhatamIdoing (talk) 08:03, 28 May 2017 (UTC)
      Depends on a rather silly quibble about coffman's use of "best", we do encourage better sourcing not weaker sourcing in multiple places including, "Base articles on reliable, third-party, published sources with a reputation for fact-checking and accuracy." . . ."Editors may also use material from reliable non-academic sources, particularly if it appears in respected mainstream publications." . . . . "In general, the more people engaged in checking facts, analyzing legal issues, and scrutinizing the writing, the more reliable the publication." . . . "When available, academic and peer-reviewed publications, scholarly monographs, and textbooks are usually the most reliable sources." . . . The opinions of specialists and recognized experts are more likely to be reliable and to reflect a significant viewpoint" . . . -- Alanscottwalker (talk) 10:29, 28 May 2017 (UTC)
      Alanscottwalker I have sampled the articles you have created and I think that you do what most editors do and use the most reliable sources you can find, and with one or two exceptions most of the citations in most of the articles you have created are online (BTW did you really need 11 citations to support three sentence in Chicago Architecture Biennial? If you do then, please consider bundling them together). I assume from your sampled edit profile that if you wanted some information in EB1911 you would not go a find a hard copy. If I am wrong please confirm that. I suspect that you would use one of the on-line copies either a facsimile used at the Internet Archive or a copy on Wikisource:EB1911. Wikipedia has a whole project devoted to linking EB1911 articles to copied of the article on Wikisource where there is a whole project dedicated to creating machine readable proofread copies of the EB1911 text. Which of the three sources are preferable to include as a citation? Should the editors who add links to the proofread text on Wikisource pack up and go home or ought they to continue to provide readers with access to machine readable copies of the EB1911 text within citations to EB1911? -- PBS (talk) 12:19, 28 May 2017 (UTC)
      What a quixotic mission you set for yourself, and how grossly weird that you would attempt to make this personal. But, yes, I have used all kinds of sources, and tried always to pay attention to their quality for the edit. But if I rely on a print source, I add the print source. As for citing many sources, I do that because I think it helpful in promoting article expansion and in promoting information for the reader - is it strictly necessary perhaps not, but so what. Alanscottwalker (talk) 15:17, 28 May 2017 (UTC)
      A long string of citations also has a useful, practical effect: it helps other editors (such as New Page Patrollers) realize that this is a notable subject, even if they've WP:NEVERHEARDOFIT or think that ITSNOTIMPORTANT. WhatamIdoing (talk) 17:20, 28 May 2017 (UTC)
  • Oppose per Finnusertop. Our ultimate goal should be to use the best sources we can, which will vary depending on the topic. For some articles, the best sources will be online stories and for others they will be books or journal articles that aren't as easily accessible. I wouldn't want to see an environment where people routinely challenge facts that are cited to print sources but cannot be verified online, and this proposal would be a step towards that. Giants2008 (Talk) 15:55, 28 May 2017 (UTC)


Arbitrary section break

Summarizing some of the discussion so far.

  1. Citations are for the reader's benefit as well as the editor's.
    "are there to verify allow readers to have confidence that the WP material they read is accurate." -- PBS
    "the service to readers is a highly reliable/reputable source confirming the content" -- Kmhkmh
  2. Some editors remove excess or duplicate cites, only including one source link per claim.
    "Too many citations at the end of a sentence decreases readability. One citation is enough for each sentence in most cases." -- QuackGuru
    [adding many extra citations] "would lead to extremely cluttered articles" --Finnusertop
  3. When a source is available in multiple locations, some free and some not, there is disagreement about which should be preferred.
    "In some cases, linking to the free version instead [or as the default] is not a good idea." --SarahSV
    Some editors prefer print sources -- the proliferation of low-quality "pay to publish" OA journals, sometimes used as sources for dubious claims, means OA-only sources are viewed with caution
    Adding additional links to more open versions inside a cite is an option, but has limitations
    The converse cite template does not exist (a cite to an open version, with an optional additional link to a closed source such as a print version).
  4. Quality and Appropriateness are the only considerations for sources
    Other editors need to be able to find the source to confirm quality and appropriateness
    "Other people... can check" already implies accessibility. The question is how much.
Questions
  • What is the value to readers of being able to confirm claims directly?
    Ex: someone cites a dubious claim to a 500-pg book, with no page numbers.
    Ex: someone cites a claim to a book long out of print, not available from most libraries.
    Ex: the default link in a cite is paywalled, with a tiny link to a free version, vs. the opposite
  • What are alternatives to finding an open copy, when using an obscure or hard to reach source?
    Long-form quotes, screenshots, multiple secondary sources making similar claims about the source material
  • What is the danger of inaccessible cites? Distinguish practice from theoretical edge cases.
    Cranks or POV pushers keeping inaccurate claims in with obfuscated cites
    Hiding the existence of accessible cites (people may remove them)
  • What is the danger of adding more-accessible cites?
    Article clutter
    Editors may stop caring about source quality (unclear how this follows...)
    Hiding the existence of inaccessible cites which may be higher quality (people may remove them)

– SJ + 22:18, 29 May 2017 (UTC)

Defining what verifiability means

I'm feeling again like some editors struggle with the difference between material that is verifiABLE and material that is conveniently CITED. (Also, I'm thinking again about those researchers a while ago, who read the first few sentences of WP:V and decided that if the source wasn't available for free on the internet, then it violated this policy.) Do you think that we should consider a new sentence in the lead, that begins with something like, "A piece of information is considered to be verifiable if..."

It might go on to say something like "if any source reliable for that information, whether cited in the article or not, has previously published that information, anywhere in the world, in any language, and that source is still available to people that are willing and able to invest the necessary time, effort, skill, and money to find, obtain, and read the source", which is a lot of words, but it might give people a better idea of how "someone is able to verify that this material could have come from a reliable source" differs from "this sentence in the article has already been cited". What do you think? WhatamIdoing (talk) 08:14, 28 May 2017 (UTC)

This discourages using inline citations and encourages unsourced content. I support indef blocks for editors continuing to add unsourced content. I want to make it part of policy for admins to block editors for adding challenged unsourced content. QuackGuru (talk) 15:02, 28 May 2017 (UTC)
I'm not sure that the suggestion wouldn't add to the confusion. The lede appears to contradict itself, starting with the first sentence:

In Wikipedia, verifiability means that other people using the encyclopedia can check that the information comes from a reliable source. ... All quotations, and any material whose verifiability has been challenged or is likely to be challenged, must include an inline citation that directly supports the material.

So verifiabily means that people can check that the info comes from a RS, but you don't have to include one if you don't think the info is going to be challenged. Not because the RS's for that information are easy to find, or the information is universally obvious, or some other reason obviating the need to be able to check it, but because the adding editor thinks that it's not going to be challenged, regardless of the experience or knowledge of the editor making that evaluation. The proposed addition would add another definition to the one in the first sentence of the lede and that's just confusing. The reason that some might think that such an additional definition might be needed is because while the term verifiability retains some of its ordinary English meaning, it's really a term of art that the entire balance of the policy defines. It's like the term reliable source, which has a meaning here related to but entirely different from its plain English meaning. Frankly, a sentence saying something like "What verifiability means for purposes of Wikipedia and how that meaning is to be implemented are defined by this policy." would seem to me to be more appropriate. It is important as well, in my opinion, to recognize — and perhaps put in the policy — that the "likely to be challenged" standard is not really a policy to determine what should and should not be cited for the purpose of fulfilling the goal set out in the first sentence of the lede, but is instead a policy to encourage editing, especially by newcomers. The pernicious effect of that policy on the goal set out in the first sentence is counterbalanced by two later elements of the policy: first, the right to challenge unsourced information without any burden other than a concern (with good faith implied) that it may be unverifiable and second, the right to express that challenge either through discussion, through tagging and discussion, or through mere removal without any additional burden except (if this much) expressing that you have that concern. The policy makes it easy to include unsourced information to encourage editing, but then also makes it easy to challenge or remove unsourced information and that's, at the end of the day, a good plan. Best regards, TransporterMan (TALK) 18:55, 28 May 2017 (UTC)
That is promoting a negative practice IMO. This is a better recommendation "any material whose verifiability has been challenged or is likely to be challenged, must include an inline citation that directly supports the material" Doc James (talk · contribs · email) 18:57, 28 May 2017 (UTC)
Not sure what you mean, Doc, or who or what you're responding to. What you quote is what the policy already says. Best regards, TransporterMan (TALK) 19:29, 28 May 2017 (UTC)
Responding to the opening comment in this thread by WAID. I think what we have is good. Doc James (talk · contribs · email) 00:14, 29 May 2017 (UTC)
I don't think that this is a negative practice at all. The rules are that:
  • you must not add anything at all(!), unless you are able to cite a published, reliable source to support it (e.g., upon request from another editor), and
  • you must not add four specified types of information (which happen, incidentally, to cover half of BLP content and most medical and scientific content), unless you are already citing a source to support it.
So if Doc James wants to add "The human hand normally has five digits" to Hand, then he need not cite a source (unless asked), because we're all pretty certain that this could be verified in any basic anatomy textbook. On the other hand, if he wants to add "Syndactyly is one of the most common birth defects involving the hands or feet", then he needs to cite a source for that promptly (ideally in the same edit). Although this is both true and verifiable (and probably even belongs in Hand), statements about epidemiology are exactly the kind of thing that is WP:LIKELY to be challenged at some point, and therefore he must only add it with a citation. WhatamIdoing (talk) 23:01, 28 May 2017 (UTC)

IMO in this area, if one understands the fuzzy nature of how Wikipedia operates, is currently pretty well written in this respect. EXCEPT, there is one area where the policy it is schizophrenic. The policy and it's dictated operational mechanics should support the idea that a challenge should include a good faith concern that the material may not be verifiable. North8000 (talk) 20:22, 28 May 2017 (UTC)

Well, I mostly agree with you, but I think it's a bit off topic. Before people can decide that they have a good-faith concern that the material may not be verifiable, we ought to have a good shared understanding about what it means for something to be verifiable. Otherwise, you end up with people talking past each other:
  • "I think it's verifiable; see, I found this source at my university research library."
  • "Well, I can't afford to buy that source, so I say it's not verifiable as far as I'm concerned".
If we define the term – e.g., "Verifiable means that someone can verify it, not that you personally can" or whatever definition we think is reasonable – then editors can't make up their own definitions of what does/doesn't count as verifiable. WhatamIdoing (talk) 23:09, 28 May 2017 (UTC)
I oppose your suggested wording in the initial post per it being too wordy and per what I argued above. Flyer22 Reborn (talk) 21:17, 30 May 2017 (UTC)